Mutational heterogeneity in low-density lipoprotein receptor gene related to familial hypercholesterolemia in Morocco

(1)

Mutational heterogeneity in low-density lipoprotein receptor gene related to familial hypercholesterolemia in Morocco

R. Chater â,b, K. Aït Chihab â,b, J.P. Rabès^c,d, M. Varret^c, L. Chabraouiê, Y. El Jahiri^f, A. Adlouni ^b, C. Boileau ^c,d, A. Kettani^b, M. El Messalâ,^⁎

aLaboratoire de Biochimie, Groupe de Génétique et Biologie Moléculaire, Faculté des Sciences Aïn Chock, BP: 5366 Maarif, Casablanca, Morocco

bLaboratoire de Recherche sur les Lipoprotéines et l'athérosclérose, Faculté des Sciences Ben M'Sik, Casablanca, Morocco

cINSERM U781, Université Paris 5, Hôpital Necker-Enfants Malades, Paris, France

dLaboratoire de Biochimie, d'hormonologie et de Génétique Moléculaire, Université Versailles-Saint-Quentin-en-Yvelines, AP-HP, Boulogne, France

eLaboratoire de Biochimie, Centre d'étude des Maladies Héréditaires du Métabolisme, Hôpital d'Enfants, Rabat, Morocco

fService de Biochimie Clinique, Hôpital Militaire Avicenne, Marrakech, Morocco Received 14 March 2006; received in revised form 1 May 2006; accepted 2 May 2006

Available online 16 May 2006

Abstract

Background:Familial hypercholesterolemia (FH) is an autosomal dominant disorder caused by mutations in the low-density lipoprotein receptor (LDLR), apolipoprotein B (APOB) and proprotein convertase subtilisin/kexin type 9 (PCSK9) genes. Until now, molecular data concerning FH in Morocco is still limited. To gain more information in this field and to assess the contribution of these three genes in the cause of FH determinism, we analyzed six unrelated Moroccan probands and twenty-five of their family's members.

Methods:AfterLDLRandAPOB genotype analysis, we screened theLDLRgene for mutations using southern blot and PCR-sequencing analysis.

We also screened theAPOBgene for the two common mutations R3500Q and R3531C by PCR-mediated site-directed mutagenesis. ThePCSK9 gene was analyzed by direct sequencing.

Results:We identified three novel mutations (C25X, IVS3+5G>T, D558A) and two mutations previously described (D151N, A480E) in theLDLR gene. The R3500Q and R3531C mutations are absent in our probands and for 1 proband, the implication ofLDLR,APOBandPCSK9genes was excluded, supporting the implication of a fourth gene in the determination of FH.

Conclusion:These data are in agreement with our previous study that suggests a heterogeneous mutational spectrum of FH in Morocco.

Keywords:Familial hypercholesterolemia;LDLR;APOB;PCSK9; Genetic heterogeneity; Moroccan families

1. Introduction

Familial hypercholesterolemia (FH) is one of the most common inherited metabolic diseases with an average world-

wide prevalence for heterozygous patients of about 1 in 500.

The homozygous form is less frequent, affecting approximately one per million. FH is an autosomal dominant disorder characterized by marked elevation of serum low-density lipoprotein cholesterol (LDL-C) levels, deposit of cholesterol in several tissues and premature coronary heart disease (CHD) [1]. The FH phenotype is caused by defective LDL clearance mainly due to mutations in LDL receptor (LDLR) gene[2]and also to defects in apolipoprotein B-100 (APOB) gene (familial Defective APO B-100 or FDB)[3], in proprotein convertase subtilisin/kexin type 9 (PCSK9) gene [4], or in other still- unidentified genes[5]. The main causes of FH are associated with mutations in the LDLR gene. Until now, over 1000

Abbreviations:APO B, apolipoprotein B; CHD, coronary heart disease;

FDB, familial defective APO B-100; FH, familial hypercholesterolemia; HDL, high-density lipoprotein; HDL-C, HDL-cholesterol; LDL, low-density lipoprotein; LDL-C, LDL-cholesterol; LDLR, LDL receptor; PCR, polymerase chain reaction; PCSK9, proprotein convertase subtilisin/kexin type 9; TC, total cholesterol; TG, triglycerides.

⁎Corresponding author. Tel.: +212 22 23 06 80/84; fax: +212 22 23 06 74.

E-mail address:elmessal@yahoo.fr(M. El Messal).

doi:10.1016/j.cca.2006.05.007

(2)

different sequence variations have been described in this gene, including nonsense and missense mutations, duplications and insertions[6]. However, the analysis of the molecular basis of FH in different ethnic groups shows a variation in prevalence of mutations in theLDLRgene. While in some populations FH is associated with a small number of LDLR gene mutations, in others the mutation spectrum of this gene is very heterogeneous.

A first report ofLDLRgene mutations in Moroccan FH patients described by our group showed a large variety of defects in this gene related to FH [7]. In our previous study, the molecular identification of mutant alleles was determined in all probands, except for two. In the first one who is a compound heterozygote, only one allele defect had been characterized. However, in the second proband, the molecular basis of FH was not identified.

In the present study, we provide further information about genetic variations in these probands. We have also enlarged our investigation on the molecular basis of FH in Morocco by analyzing the implication ofLDLR,APOBandPCSK9genes in a novel group of probands. Due to the high incidence of premature cardiovascular disease in FH population and to obtain as soon as possible an early diagnosis, it is mandatory to search for FH patients in proband's families, especially before the clinical manifestation of this pathology[8]. For this purpose, we have recruited first- and second-degree relatives of FH index cases in this study.

2. Methods

2.1. Patients

Twenty-four subjects from four unrelated Moroccan families were analyzed, including four probands and twenty first- and second-degree members of their families. Two other unrelated probands, previously described by our group [7] and five of their relatives have also been included in this study. In total,

thirty-one individuals were studied. Diagnosis criteria for FH proband were high plasma levels of total and LDL cholesterol, family history of hypercholesterolemia especially in children, presence of cholesterol extra-vascular deposits and personal and/or family history of premature CHD [2]. The proband's relatives were diagnosed for FH according to the MEDPED criteria [9]. These criteria were best defined in the proband's relatives because the age and the degree of familial relationship with the proband were taken into account. Secondary cause of hypercholesterolemia, including diabetes, hypothyroidism and nephritic syndrome were excluded. Demographic data and detailed family history questionnaire and a complete medical history were obtained for all subjects. A control cohort is made of 107 unrelated healthy Moroccan individuals with no family history of hypercholesterolemia. None of the subjects were under medications that might affect lipid metabolism. Blood samples were collected from all participants in the study after their informed consent.

2.2. Biochemical analysis

Blood samples were collected after 12 h fasting period and immediately centrifuged at 4 °C for 10 min at 3000 rpm. Total cholesterol (TC), triglyceride (TG) and high-density lipoprotein cholesterol (HDL-C) were measured using enzymatic methods (Roche Diagnostics). The HDL-C fraction was obtained after precipitation from plasma of APO B-containing lipoproteins with phosphotungstic acid and MgCl2. LDL-C was calculated by the Friedewald formula[10].

2.3. Genetic analysis 2.3.1. DNA extraction

Genomic DNA was isolated from frozen whole blood EDTA samples using the salting-out method described previously[11].

Table 1

PCR oligonucleotide primers for theLDLRgene mutation analysis

Exon no. Forward primer sequence (5′to 3′) Reverse primer sequence (5′to 3′) PCR product (base pair) Annealing temperature (°C)

Promoter TTGAAAGGCTGTTGTTATCCTTCTG ACGACCTGCTGTGTCCTAGCTG 396 62

Exon 1 TCCTCCTCTTGCAGTGAGGTGAAG TCCCTCTCAACCTATTCTGGCG 298 59

Exon 2 GTTGTGCTTGCTTAATTCCCTG CAACATGGCGAGACCCTGTC 339 59

Exon 3 GGTCTTGAACCCCTGACCTCAC AACACTCCCCAGGACTCAGATAG 311 59

Exon 4.1 ATGAGGAAACTGAGGCACCGAG TGTCCCCTTGGAACACGTAAAG 443 63

Exon 4.2 TCCAGTGCAACAGCTCCACC GTTGGAAATCCACTTCGGCAC 341 63

Exon 5 GAACTCCTGGGCTCAAGCAATC GTGAGGCTCTGAGAAGTCAAGTCAC 420 64

Exon 6 TTACAGGCACAAACCACCGTG ACCCTACAGCACTCATGTCTCAGTC 350 57

Exon 7 GGTGGAGGTTGTAATGAGCCAAG TGGTTGCCATGTCAGGAAGC 267 60

Exon 8 GTTTCCTTGATTACATCTCCCGAG GAGTCTGTGCAAAGTTCAGAGGATG 359 55

Exon 9 AAAGTGCTGGGATTACAGGTGGGC TGGATGTCTCTGCTGATGACGTGG 509 64

Exon 10 TGACCTGTCCCAGAGAATGATCTG CCCACTAACCAGTTCCTGAAGCTC 415 55

Exon 11 TTCGTTGCACGCATTGGC GGGAAACCTTCAGGGAGCAG 300 59

Exon 12 AGGTGCTTTTCTGCTAGGTCCCTG TGCGTTCATCTTGGCTTGAGTG 343 59

Exon 13 GGCAACCCCCGTGAAACTCTGTCTG GCAGGAACGAGATCATCAGCTATTC 393 55

Exon 14 CTTGAAACCTCCTTGTGGAAACTC TGACAGATGAGCAGAGAGAGGCTC 335 59

Exon 15 TGGTATTTTGCCATGTTGACCAG AGGACGACACCTGGACTCCATC 439 59

Exon 16 TGGGAAGTTCTCCAAGTGTCCAG TCACATAGCGGGAGGCTGTG 262 57

Exon 17 GGCGATCTCTAAACAAACATAAAAG TGTCCTCGATCTGGAGGGC 449 55

Exon 18 ACTCACCGTCTCCCTCTGGC AAAGGAAGAAACCAAAATCCCAAC 320 57

(3)

2.3.2. Genotyping of LDLR and APOB genes

For LDLR gene, genotyping and co-segregation analysis were carried out using four polymorphic microsatellites markers, two intragenic markers (D19S584 in intron 1 and the (TA)n in exon 18) and two flanking markers (D19S394 and D19S221). The PCR amplification of these markers was carried out according the method described by Amasino[12].

The haplotype co-segregating with the LDLR gene mutation was identified by analysis of these four polymorphic microsatellites.

Concerning the APOB gene, the haplotypes of FH-family members were determined by two polymorphic markers, the 5′

HVR (TG repeat) and 3′HVR (VNTR).

2.3.3. Analysis of LDLR and APOB genes

The promoter, 18 exons and their flanking sequences of the LDLRgene were amplified by polymerase chain reaction (PCR).

The PCR reaction mixture of 25 μL contained 50 ng DNA, 100μmol/L of each dNTP, 15 pmol of each primer, 2.0 mmol/L MgCl2, and 0.5 U ofTaqDNA polymerase (GIBCOBRL). The

Table 2

Clinical and biochemical data of FH probands and their relatives

Subjects Sex Age TC TG LDL-C HDL-C CEVD CHD Clin. Dx

Family BA

1BA M 51 329 89 257 54 − + Heterozygous FH

2BA M 35 391 109 323 46 − − Heterozygous FH

3BA^a(proband) M 47 320 132 231 62 − − Heterozygous FH

4BA F 44 237 70 183 40 − − Not affected

5BA M 42 145 101 110 65 − − Not affected

Family K

10M^a(proband) F 18 600 131 536 38 + +¹ Homozygous FH

11M M 38 360 110 341 35 − − Heterozygous FH

Family A

1A (proband) M 55 674 120 622 28 +^b +^{2, c} Homozygous FH

2A F 44 185 86 126 42 − − Not affected

3A M 19 258 137 182 49 − − Heterozygous FH

4A F 17 291 58 238 41 − − Heterozygous FH

5A F 15 332 109 260 50 − − Heterozygous FH

Family BO

1BO M 62 303 99 240 43 − − Heterozygous FH

2BO F 47 511 91 454 39 − − Heterozygous FH

3BO M 26 167 90 114 35 − − Not affected

4BO M 24 230 105 169 40 − − Not affected

5BO F 21 133 51 79 44 − − Not affected

6BO M 18 287 100 222 45 − − Heterozygous FH

7BO M 14 212 112 146 44 − − Not affected

8BO (proband) M 12 739 79 703 20 + +¹ Homozygous FH

Family FO

1FO M 70 263 63 218 33 − − Heterozygous FH

2FO F 57 305 72 246 44 − − Heterozygous FH

3FO F 34 150 54 101 38 − − Not affected

4FO (proband) M 26 596 65 551 33 + +³ Homozygous FH

5FO M 20 121 68 76 31 − − Not affected

6FO M 15 125 58 69 44 − − Not affected

Family Y

1Y F 65 292 60 232 48 − − Heterozygous FH

3Y (proband) F 46 304 80 247 41 − +³ Heterozygous FH

5Y M 37 151 136 90 34 − − Not affected

M: male, F: female, TC: total cholesterol, TG: triglycerides, LDL-C: LDL-cholesterol, HDL-C: HDL-cholesterol, CEVD: cholesterol extra-vascular deposits, CHD:

coronary heart disease (¹artery injuries,²myocardial infraction and two coronary bypasses,³stable angina and dyspnea), Clin. Dx: clinical diagnosis.

Lipid values are in mg/dl and represent values obtained before any medical treatment.

Clinical diagnosis of proband's relatives according to Williams et al.[9].

a Previously reported in Ref.[7].

b First CEVD at 30 years.

c First coronary manifestation at 35 years.

(4)

PCR conditions of the total 30 cycles performed in a thermal cycler (Perkin Elmer) were 94 °C/30 s, annealing temperature/

30 s and 72 °C/10 s that were preceded by a 2 min denaturation at 94 °C and followed by a 5 min extension at 72 °C. The primer sequences and annealing temperature used for each amplified fragment are shown inTable 1.

Direct sequence analysis of purified PCR product (Sepha- dex G50 and P100 spins columns, Biorad) was carried out using the BigDye™ terminator cycle sequencing ready reaction kit on ABI PrismTM 3100 DNA sequencer (Applied Biosystems). When a mutation was detected, another sequencing reaction was performed both on genomic DNA

Fig. 1. Pedigrees of families A (A), BO (B), FO (C) and Y (D) showing lipid values,LDLRgene mutations and haplotypes. The probands are indicated by an arrow.

Unaffected individuals are indicated with open symbols; those affected are shown with half-filled symbols (FH heterozygote) or completely filled symbols (FH homozygote). Affected and unaffected subjects are considered in this figure according to presence or absence ofLDLRgene mutations. Symbols with a slash denote deceased individuals. The total cholesterol (TC) and LDL-cholesterol (LDL-C) concentrations are given in mg/dl. Age is given at lipid measurement. Haplotypes were constructed with polymorphic microsatellite markers that are ordered from 5′to 3′.

(5)

from a relative and from a new PCR product from the proband.

To screen large rearrangement in theLDLRgene, southern blot analysis was performed as previously described by our group[7].

Concerning the APOB gene, the two common mutations R3500Q and R3531C were simultaneously screened in all subjects using the PCR-mediated mutagenesis method and double restriction of a unique PCR product, as previously described[13].

2.3.4. Confirmation of nucleotide changes

Restriction endonuclease digestion of PCR amplicons was used to confirm the IVS3+5G>T and D558A mutations.

Concerning IVS3+5G>T mutation, amplification was carried out by PCR-mediated site directed mutagenesis using a modified PCR primer (5′AGACGAGCAAGGCTGTCCTAA3′

and 5′CGGAAGAGGCTTGGTATGAG3′, withCthe changed nucleotide at position 28502 or 313+1). The modified forward primer introduces aDdeI restriction site in the normal sequence that is not generated when the mutation is present. Whereas, D558A directly abolishes a restriction endonuclease site for HinfI. The samples were run on 2% or 3% agarose melting gel and visualised on a UV transilluminator after staining with ethidium bromide. To confirm the presence of D151N and A480E mutations, we used SSCP analysis. Briefly, 5μL of the PCR product was added to 10μL of loading buffer (formamide, EDTA, bromophenol blue and xylene cyanol), denatured at 96

°C for 5 min, and immediately placed on ice. Electrophoresis was carried out in a vertical 10% or 12% polyacrylamide gel at 100 Vovernight, using two different temperatures (10 °C and 20

°C). DNA bands were detected by silver nitrate staining revelation.

2.3.5. Analysis of PCSK9 gene

PCSK9 gene was screened by direct sequencing of its 12 exons and their flanking regions as previously described[14].

3. Results

3.1. Clinical analysis

Biochemical and clinical data of six unrelated probands (four homozygotes with parental consanguinity and two heterozy-

gotes) and their twenty-five relatives are shown in Table 2.

Among the proband's relatives, we identified fourteen novel heterozygous FH patients. Cholesterol extra-vascular deposits were observed only in the homozygous FH and CHD was reported those probands and in one heterozygous FH.

3.2. Genetic analysis

Following genotyping of the APOB locus and DNA screening for the R3500Q and R3531C mutations, FDB was excluded in any of the studied subjects.

The genotyping of LDLR gene for the four polymorphic markers was possible in five families. In families A, BO, FO and Y, co-segregation was observed showing an association between an LDLR allele with FH phenotype (Fig. 1).

However, the implication of theLDLR gene was excluded in family BA.

Southern blot analysis of genomic DNA from all probands as well as one of their family members showed no majorLDLR rearrangement, suggesting the presence of point mutations or minor rearrangement. After sequencing all exons, including the intron–exon boundaries and the promoter, fiveLDLRmutations were detected: one nonsense mutation (C25X) in exon 2, one splice mutation (IVS3+5G>T) in intron 3 and three missense mutations in exons 4 (D151N), 10 (A480E) and 12 (D558A) (Table 3). Among these mutations, C25X, IVS3+5G>T and D558A are reported here for the first time. The remaining two mutations D151N and A480E were previously described in Danish [15] and Japanese populations [16], respectively. All these five mutations were absent in 107 normocholesterolemic control subjects.

–The C25X mutation causes a premature stop codon at cysteine 25 giving rise to a truncated LDL receptor (data not shown). This mutation was found in heterozygous state in the 2 members of family K: in a compound heterozygote carrying also P664L mutation previously reported[7]and in his heterozygous brother.

–The mutation IVS3+5G>T was identified in homozygous and heterozygous states in proband 1A and all his children, respectively. This transversion was not found in 2A (proband's wife) who was normo-cholesterolemic. This result was confirmed by haplotype analysis of the LDLR locus (Fig. 1A).DdeI RFLP analysis was used to confirm the

Table 3

LDLRgene mutations identified in 5 Moroccan FH probands

Exon Mutation name^a Nucleotide change^b Confirmation Number of patients carrying the mutation Reference

2 p.C25X c.138C>A Second sequencing 1 compound heterozygous, 1 heterozygous This report

3–4 IVS3+5G>T 313+5G>T DdeI⁻ 1 homozygous, 3 heterozygotes This report

4 p.D151N c.514G>A SSCP 4 heterozygotes [15]

10 p.A480E c.1502C>A SSCP 1 homozygous, 2 heterozygotes [16]

12 p.D558A c.1736A>C HinfI⁻ 1 homozygous, 5 heterozygotes This report

−: Mutation destroys restriction site (for confirmation of IVS3+5 mutation, amplification was carried out by PCR-mediated site directed mutagenesis using a modified PCR primer, see Methods).

a Named according to the amino acid numbering of Yamamoto et al.[26].

b Nomenclature at the DNA level, according to international nomenclature and the nomenclature working group[6].

(6)

presence of the mutation in the proband and the family members. No other mutation was found in this family.

– The mutation D558A changes an aspartate to alanine in the fifth of the six YWTD repeats in the EGF precursor homology domain[17]. The proband 8BO was homozygous for this mutation. This mutation was confirmed by HinfI RFLP analysis. The consanguineous proband's parents and three of their siblings were heterozygous for the D558A mutation.Fig. 1B shows the segregation of this mutation in the proband's family. Subjects 4BO and 7BO showed no correlation betweenLDLR genotype (LDLR gene mutation and polymorphic microsatellites markers) and FH phenotype (TC and LDL-C values). They are D558A carriers, but have TC and LDL-C in normal range.

– Regarding A480E mutation, the proband and his consanguineous parents are carriers of this mutation at homozygous and heterozygous states, respectively. Also, genotyping of 4 polymorphic markers for theLDLRgene (Fig. 1C) showed that the proband is homozygous and the parents are heterozygous for all markers analyzed and no other family member was affected. No otherLDLRmutation was found in this family. For the previously reported D151N mutation, heterozygosity was identified in proband 3Y and three hypercholesterolemic relatives. The co-segregation was also confirmed by genotyping analysis (Fig. 1D).

As in our previous report, the A370T variation was detected (33% of analyzed subjects in this report) and was not associated with elevated levels of LDL-C. In addition, four silent mutations were detected. They include 2 novel variations, R450R and R723R, and two previously reported in our population, N570N and V632V[7].

Regarding family BA, haplotype analysis using highly polymorphic markers ofLDLR andAPOBgenes excluded the implication of these genes in FH phenotype. Although after screening of the common R3500Q and R3531C mutations in the APOB gene and direct sequencing of all exons (and their flanking regions) of theLDLRandPSCK9genes, no mutation was detected. Studies are underway to determine whether the hypercholesterolemia in this family is caused by another locus causing FH phenotype.

4. Discussion

In Morocco, specialized clinics for dyslipidemia do not exist and management of this major cardiovascular risk factor remains inadequate, making the FH identification and management difficult if not impossible. A surprising finding from our study is the elevated number of recruited homozygotes in our sample. This is due to the high prevalence of the consanguinity in our population and to the early clinical manifestation of homozygous phenotype, essentially the very large and spectac- ular cholesterol extra-vascular deposits which facilitate the homozygous FH diagnosis. However, the diagnosis of heterozygous FH is more difficult. Generally, these patients are referred to a cardiologist when cardiovascular complications appear and the disease's etiology is rarely investigated.

All data concerning this pathology in Morocco originate from our group. Unfortunately, the number of recruited unrelated FH subjects remains insufficient for the prevalence assessment of the FH causing mutations. However, the identification of FH mutations in a low number of unrelated FH probands can provide us some data to speculate about the mutational spectrum related to FH in the population. A first report of LDLR gene mutations in Moroccan FH patients has been described recently by our group[7]. In this previous study, using PCR/SSCP/sequencing, the molecular identification of mutant alleles remained unknown in two probands. By using other approaches, we have detected a novel mutation, C25X, in a compound heterozygote proband (10M) and his heterozygous brother (11M). Also, we have excluded the implication of the LDLR,APOBandPCSK9genes in FH in the proband 3BA and his relatives. This last result constitutes an additional confirmation to the genetic heterogeneity in FH and suggests the existence of additional FH gene loci.

The recruitment of four novel probands and their relatives in this report has revealed the presence of four point mutations.

In total, we have identified five mutations, three of which are reported here for the first time.

The C25X mutation creates a premature termination codon at cysteine 25 resulting in a truncated LDLR protein. Accordingly, the C25X has to be classified as a null-allele[18]and would be singularly pathogenic of FH.

We identified a novel splice site mutation, IVS3+5G>T, changing the 5th nucleotide of the 5′ splice donor site of intron 3 from G to T. This base substitution alters the highly conserved G nucleotide which forms part of the donor splice site. In fact, in 86% of vertebrate genes, including LDLR gene, the G nucleotide is present in the +5 nucleotide position of the donor splice site[19]. At the same nucleotide, the IVS3 +5G>A transition was previously described by Liguori et al.

[20]. This mutation is known to cause an abnormal splicing event by causing a deletion of exon 3 and therefore leading to the pathogenesis of FH [20]. Accordingly, the IVS3+5G>T mutation identified in this study would affect the mRNA splicing; however mRNA studies must be performed to confirm it and to characterize the type of abnormal splicing event. The pathogenicity of IVS3+5G>T mutation is also supported by mutation co-segregation in hypercholesterolemic subjects and its absence in the normolipidic relatives as well as by its absence in the 107 non-FH control subjects. In spite of these facts supporting the FH causal effect of IVS3+5G>T mutation, it is surprising to note the delay in the onset of clinical manifestations (Table 2) and the considerable increase in the life expectancy observed in the homozygous proband 1A, who is 55 years old. This could be related to the nature of the IVS3+5G>T mutation, which still has to be determined. It also could be due to some other genetic and/or environmental factors.

The D558A mutation occurs in the fifth of the six YWTD repeats in the EGF precursor homology domain that are involved in receptor recycling and lipoprotein release at low pH. Of the mutations causing FH, 54% are located in the EGF precursor homology region of theLDLRgene and 34% among

(7)

the YWTD repeats[17]. Moreover, the aspartate (D) changed in D558A mutation is a conserved amino acid in each YWTD motif, which seems to play a crucial role in defining and stabilizing the propeller structure[17]. In addition, aspartate at position 558 is strongly conserved in LDLR protein of multiple species including mouse, rabbit, hamster and rat, compared to the human LDLR protein sequence. Two other mutations at the same position, D558Y and D558N, have been associated with FH phenotype[18,21,22]. Therefore, substitution of negatively charged aspartate by neutral alanine would be expected to disturb normal receptor function and to cause FH. Indeed, this mutation was found at homozygous and heterozygous states in the proband and his consanguineous parents, respectively, and it was absent in normolipidic control subjects. However, we have noted that two D558A heterozygous carriers have normal TC and LDL-C levels, which purports genetic testing to improve FH diagnostic capacity, especially in young FH patients who may present overlapping values with unaffected individuals.

In our sample, we also identified two mutations (D151N and A480E), previously reported only once in other ethnic groups, the Danish [15] and Japanese [16], respectively. The D151N mutation is located in the fourth repeat of the binding domain.

Jeon and Blacklow[23]reported that the fourth repeat is, with the fifth repeat, the key ligand-binding domain so that the receptor adopts a closed conformation and direct contact with the top face of the beta-propeller domain at low pH. D151N affects the highly conserved aspartate that is part of the clusters of conserved acidic residues with the signature sequence DCxDxSDE (where x is not a conserved amino acid) located just before the C-terminal of all ligand binding repeats [23], supporting causal effect of this mutation. A480E missense mutation causes a change of neutral alanine to negatively charged glutamate in the beta-propeller domain of the LDL receptor. This mutation would result in a partially defective transport (class 2B)[16]of the LDL receptor. Moreover, the co- segregation of D151N and A480E with the hypercholesterolemic phenotype in families Y and FO, the absence of these mutations in non-affected relatives and in 107 control subjects, provided additional evidence of the pathogenicity of these mutations.

Identification of A370T in 33% of analyzed subjects, the non-co-segregation of this sequence variation with the FH phenotype and its presence in families in which the D151N and A480E mutations segregated demonstrate the common charac- ter of this polymorphism. Furthermore, these findings are corroborated by other studies[7,24,25].

In all Moroccan FH subjects analyzed so far (current study and Ref.[7]), theAPOBgene defects appear to be absent and the heterogeneity ofLDLRgene mutations obvious (one proband/

one mutation). The identification of a family with non-LDLR, non-APOBgene defects could contribute to the identification of new genes related to autosomal dominant FH, such asHCHOLA 4 and HCHOLA5. This study has allowed us to identify an important number of novel FH patients among the proband relatives, before any cardiovascular manifestation. The chal- lenge in FH patients is to prevent the appearance of premature coronary atherosclerosis and its complications. For this purpose,

early identification and adequate management and treatment are necessary. Knowledge of FH-causing mutations greatly con- tributes to unambiguous and premature diagnosis, which leads to an improved management of this disease. In this respect, two heterozygous FH were misdiagnosed by clinical diagnosis, changing the number of newly identified FH relatives from 14 to 16. Finally, this study supports a probable wide variety ofLDLR gene defects in Moroccan population. However, the sample is too small to draw any conclusion about the prevalence of genetic defects inLDLRgene in the Moroccan population as a whole.

Screening for causal defects in a large number of clinically diagnosed FH patients is still required.

Acknowledgements

This work was supported by CNCPRST/INSERM 2001- 2002, Action Integrée Maroco-Française AI-MA-107 and Programme PROTARS 2 P12/11.

We wish to thank Prof. Sellama Nadifi, Director of laboratoire de génétique humaine de la faculté de médecine de Casablanca for support and encouragement. We also thank Dr.

Said Talbi and Amy Hamilton from the University of California, San Francisco, for their critical reading of this manuscript.

References

[1] Goldstein JL, Hobbs HH, Brown MS. Familial hypercholesterolemia. In:

Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic basis of inherited diseases. 7th ed. New York: McGraw-Hill; 1995. p. 1981–2030.

[2] Goldstein JL, Hobbs HH, Brown MS. Familial hypercholesterolemia.

In: Scriver CR, Beaudet AL, Sly WS, Valle I, editors. The metabolic basis of inherited disease, 2001. 8th ed. New York: McGraw-Hill; 2001.

p. 2863–913.

[3] Innerarity TL, Weisgraber KH, Arnold KS, Mahley RW, Krauss RM, Vega GL, et al. Familial defective apolipoprotein B-100: low density lipoproteins with abnormal receptor binding. Proc Natl Acad Sci U S A 1987;84:6919–23.

[4] Abifadel M, Varret M, Rabes JP, Allard D, Ouguerram K, Devillers M, et al. Mutations in PCSK9 cause autosomal dominant hypercholesterolemia. Nat Genet 2003;34:154–6.

[5] Damgaard D, Jensen JM, Larsen ML, Soerensen VR, Jensen HK, Gregersen N, et al. No genetic linkage or molecular evidence for involvement of the PCSK9, ARH or CYP7A1 genes in the familial hypercholesterolemia phenotype in a sample of Danish families without pathogenic mutations in the LDL receptor and apoB genes. Atherosclerosis 2004;177:415–22.

[6] Villeger L, Abifadel M, Allard D, Rabes JP, Thiart R, Kotze MJ, et al. The UMD-LDLR database: additions to the software and 490 new entries to the database. Hum Mutat 2002;20:81–7.

[7] El Messal M, Ait Chihab K, Chater R, Vallve JC, Bennis F, Hafidi A, et al.

Familial hypercholesterolemia in Morocco: first report of mutations in the LDL receptor gene. J Hum Genet 2003;48:199–203.

[8] Marks D, Wonderling D, Thorogood M, Lambert H, Humphries E, Neil AW. Cost effectiveness analysis of different approaches of screening for familial hypercholesterolemia. BMJ 2002;324:1303–8.

[9] Williams RR, Hunt SC, Schumacher MC, Hegele RA, Leppert MF, Ludwig EH, et al. Diagnosing heterozygous familial hypercholesterolemia using new practical criteria validated by molecular genetics. Am J Cardiol 1993;72:171–6.

[10] Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem 1972;18:499–502.

(8)

[11] Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA from nucleated cells. Nucleic Acids Res 1988;16:1215.

[12] Amasino RM. Acceleration of nucleic acid hybridization rate by polyethylene glycol. Anal Biochem 1986;152:304–7.

[13] Rabès JP, Varret M, Saint-Jore B, Erlich D, Jondeau G, Krempf M, et al.

Familial ligand-defective apolipoprotein B-100: simultaneous detection of the Arg35003Gln and Arg35313Cys mutations in a French population.

Hum Mutat 1997; 10:160–3.

[14] Allard D, Amsellem S, Abifadel M, Trillard M, Devillers M, Luc G, et al.

Novel mutations of the PCSK9 gene cause variable phenotype of autosomal dominant hypercholesterolemia. Hum Mutat 2005;26:497.

[15] Jensen HK, Jensen LG, Meinertz H, Hansen PS, Gregersen N, Faergeman O. Spectrum of LDL receptor gene mutations in Denmark: implications for molecular diagnostic strategy in heterozygous familial hypercholesterolemia. Atherosclerosis 1999;146:337–44.

[16] Maruyama T, Yamashita S, Matsuzawa Y, Bujo H, Takahashi K, Saito Y, et al. Mutations in Japanese subjects with primary hyperlipidemia— results from the Research Committee of the Ministry of Health and Welfare of Japan since 1996. J Atheroscler Thromb 2004;11:131–45.

[17] Jeon H, Meng W, Takagi J, Eck MJ, Springer TA, Blacklow SC.

Implications for familial hypercholesterolemia from the structure of the LDL receptor YWTD-EGF domain pair. Nat Struct Biol 2001;8:

499–504.

[18] Hobbs HH, Brown MS, Goldstein JL. Molecular genetics of the LDL receptor gene in familial hypercholesterolemia. Hum Mutat 1992;

1:445–66.

[19] Krawczak M, Reiss J, Cooper DN. The mutational spectrum of single base- pair substitutions in mRNA splice junctions of human genes: causes and consequences. Hum Genet 1992;90:41–54.

[20] Liguori R, Bianco AM, Argiriou A, Paucillo P, Giannino A, Rubba P, et al.

LDL receptor cDNA sequence analysis in familial hypercholesterolemia patients: 5 novel mutations with high prevalence in families originating from southern Italy. Hum Mutat 2001;17:433.

[21] Bertolini S, Cantafora A, Averna M, Cortese C, Motti C, Martini S, et al.

Clinical expression of familial hypercholesterolemia in clusters of mutations of the LDL receptor gene that cause a receptor-defective or receptor-negative phenotype. Arterioscler Thromb Vasc Biol 2000;20:

E41–52.

[22] Bertolini S, Pisciotta L, Di Scala L, Langheim S, Bellocchio A, Masturzo P, et al. Genetic polymorphisms affecting the phenotypic expression of familial hypercholesterolemia. Atherosclerosis 2004;174:57–65.

[23] Jeon H, Blacklow SC. Structure and physiologic function of the low- density lipoprotein receptor. Annu Rev Biochem 2005;74:535–62.

[24] Kotze MJ, Langenhoven E, Warnich L, Marx MP, Retiel AE. Molecular characterization of a low-frequency mutation in exon 8 of human low- density lipoprotein receptor gene. S Afr Med J 1989;76:402–5.

[25] Weiss N, Binder G, Keller C. Heterozygosity for the missense mutation Ala370→Thr in exon 8 of the low density lipoprotein receptor gene does not cause hypercholesterolemia. Eur J Med Res 1998;3:20–4.

[26] Yamamoto T, Davis CG, Brown MS, Scheider WJ, Casey ML, Goldstein JL, et al. The human LDL receptor: a cysteine-rich protein with multiple Alu sequences in its mRNA. Cell 1984;39:27–38.