- Email: [email protected]
Allelic drop-out in the LDLR gene affects mutation detection in familial hypercholesterolemia
Available online at www.sciencedirect.com
Clinical Biochemistry 41 (2008) 38 – 40
Allelic drop-out in the LDLR gene affects mutation detection in familial hypercholesterolemia Eleftheria Laios a,⁎, Kyriaki Glynou b a
Unit of Metabolic Diseases, Choremio Research Laboratory, University of Athens, 1st Department of Pediatrics, “Aghia Sophia” Children’s Hospital, Athens 11527, Greece b Medicon Hellas S.A., Gerakas 15344, Greece Received 4 May 2007; received in revised form 19 August 2007; accepted 28 September 2007 Available online 11 October 2007
Abstract Objectives: Familial hypercholesterolemia is a monogenic disorder caused by mutations in the LDL receptor (LDLR) gene. We observed allelic drop-out during LDLR genotyping and aimed at redesigning mutation detection. Design and methods: The NanoChip microelectronic array technology and PCR restriction fragment length polymorphism analysis were used. Results: Allele drop-out caused false homozygous diagnoses and was overcome using PCR primers without polymorphisms in the primer binding site. Conclusions: This report presents the importance of allele drop-out in LDLR genotyping. © 2007 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. Keywords: Low-density lipoprotein receptor; Hypercholesterolemia; Mutation; Greece
Introduction
Materials and methods
Familial hypercholesterolemia (FH) is a monogenic autosomal dominant disorder characterized by elevated low-density lipoprotein (LDL) cholesterol, tendon xanthomas, and premature coronary heart disease [1]. FH is caused by mutations in the LDL receptor (LDLR) gene [1]. FH clinical diagnosis is based on family history and elevated LDL cholesterol. Extensive genetic screening for diagnosing FH patients is in effect in several countries. Recently, we genotyped four Greek FH patients as Greece II homozygotes. However, LDL cholesterol levels of these patients indicated a heterozygote phenotype. We hypothesized that allele drop-out had caused false homozygous diagnoses in the aforementioned cases.
Genotyping at nucleotide 858 of LDLR exon 6
⁎ Corresponding author. Fax: +30 210 7709415. E-mail address: [email protected] (E. Laios).
Blood collections were performed with informed oral consent of the participants with a clinical diagnosis of FH. Genomic DNA was isolated by use of the QIAamp DNA Blood Mini Kit (Qiagen). For detection of the Greece II mutation (C to A transition at nucleotide 858 of exon 6), DNA was amplified with the forward primer 858U (5′ biotin-CTCTGCGAGGGACCCAAC 3′) and the reverse primer 858L (5′ TTTCCTGGCTGGGGACAA 3′). The PCR product was used for NanoChip analysis according to [2]. Primer 858L resulted in allele drop-out (due to the underlined nucleotide) and was subsequently replaced by primers 858L-2 or 858L-3 which do not bind to regions containing single nucleotide polymorphisms (SNPs). Primer 858L-2 is located 56 bases downstream of 858L and primer 858L-3 is located 386 bases downstream of 858L. For PCR-RFLP analysis, DNAwas amplified with 858U and 858L-2 (5′ CGCCCAGCATCGCTTCATTTTTT 3′). The PCR product
0009-9120/$ - see front matter © 2007 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.clinbiochem.2007.09.017
E. Laios, K. Glynou / Clinical Biochemistry 41 (2008) 38–40
was digested with MnlI (NEB), followed by separation of digestion products by agarose gel electrophoresis. Greece II introduces a MnlI restriction site to alleles with the C858A mutation. To repeat NanoChip analysis of samples previously genotyped as wild-type at position 858, DNAwas amplified with 858U and 858L-3 (5′ CACAGGGTGGGCAGAGTG 3′) followed by mutation detection according to [2]. Genotyping SNP rs13306513 at nucleotide 976 of LDLR intron 6 The SNP rs13306513 (A or G allele) is located at nucleotide 976 in intron 6 of the LDLR gene. We genotyped the SNP both in the general Greek population and in FH patients carrying the Greece II mutation (C858A). For the general population, saliva samples were collected after informed oral consent of 65 unrelated Greeks. Saliva collection and genomic DNA extraction were performed with the Oragene kit (DNAGenotek). Genomic DNA of FH patients was obtained as described in the Genotyping at nucleotide 858 of LDLR exon 6 section. For PCRRFLP analysis, DNA was amplified with 858U and 858L-2 (5′ CGCCCAGCATCGCTTCATTTTTT 3′). The PCR product was digested with MnlI (NEB), followed by separation of digestion products by agarose gel electrophoresis. The A allele of the SNP introduces a MnlI restriction site. Results Using amplification primers 858U and 858L, we identified four Greece II homozygous FH cases based on the NanoChip assay [2]. We confirmed these results by sequencing the same amplicons with 858L. Sequencing revealed a homozygous genotype in all four cases. However, the clinical phenotype of these patients was not consistent with the molecular results. Investigation of the false homozygous diagnosis resulted in identification of a SNP within the primer binding region of the reverse primer (858L). The SNP (rs13306513 at nucleotide 976; A or G allele) was located at the sixth position from the 3′ end of the primer and resulted in drop-out of the A976 allele. The SNP had not been reported at the time of our experimental design. We redesigned the reverse primer and performed PCR-RFLP analysis of the four pseudo-homozygotes. Using a reverse primer without a SNP in the primer binding region (primer 858L-2), the samples were genotyped as Greece II heterozygotes. According to RFLP, in all four cases, the allele with the Greece II mutation (A858) was G at 976, while the normal C858 allele was A at 976. This is consistent with the allele drop-out (ADO) observation in the four pseudo-homozygotes. The four patients in which ADO was observed belonged to three different families. We have genotyped an additional 33 unrelated Greece II heterozygotes at position 976 and have found no A alleles. Using the redesigned reverse primer 858L-3 and NanoChip analysis, we also repeated genotyping of 50 samples which we had previously characterized as normal at position 858 of exon 6 using PCR primer 858L [2]. Based on the repeated NanoChip analysis, the 50 samples appeared as wild type. This indicated that the samples had not been falsely genotyped as normals (ADO due
39
to A976 on an A858 allele could have resulted in a Greece II heterozygote falsely appearing as a normal). The prevalence of the SNP (rs13306513; A or G allele) at position 976 in the general Greek population was estimated by studying a total of 130 alleles from 65 unrelated individuals. We identified 125 G alleles and 5 A alleles. The allele frequency was estimated as 0.96 for the G allele and 0.04 for the A allele. Genotype frequencies were 0.96 for GG, 0.04 for AG, while AA was not found. Linkage disequilibrium (LD) between the two SNPs was estimated in a population consisting of the 36 unrelated heterozygote C858A patients and the 65 unrelated individuals from the general population (D′ = 0.031, r2 = 0). However, it is difficult to conclude that the two SNPs are not co-segregating since the LOD score associated with the LD calculations was very low (LOD = 0). This is probably due to the very low frequency of the rare allele at position 976 and depicts an inherent problem in LD calculation using the metrics of D′ and r2 when allele frequencies at the studied SNPs are skewed. Discussion Our results show that the presence of a polymorphism in the primer binding site leads to allelle drop-out in exon 6 of the LDLR gene, creating a potential problem in FH molecular diagnosis. ADO caused four FH heterozygotes to be falsely genotyped as homozygotes. In general, misgenotyping of an FH patient with a clinical heterozygote phenotype as a homozygote with two mutant alleles is easily detected since this would require a severe clinical phenotype and two heterozygous FH parents. However, misgenotyping of an FH patient with a clinical heterozygote phenotype as a normal with two wild-type alleles would lead to unnecessary analysis of the remaining LDLR gene for mutation identification. In addition to the LDLR gene, the apolipoprotein B100 gene and the proprotein convertase subtilisin/kexin type 9 (PCSK9) gene are known to cause the same autosomal dominant hypercholesterolemic phenotype. In most populations, only about 70–80% of heterozygous FH phenotypes are explained by mutations in the LDLR and apolipoprotein B100 genes, while PCSK9 accounts for a small percentage of the remaining nonLDLR/non-apolipoprotein B100 cases [3,4]. Although it is possible that additional genes are responsible for the autosomal dominant hypercholesterolemia phenotype, it is also likely that diagnoses may be missed due to ADO. Several studies have shown that polymorphisms within the primer binding site can result in ADO due to preferential amplification of one allele [5–8]. Therefore, primer design should include checking more than one SNP database for polymorphisms in the binding site. Since these databases are frequently updated, they should be continuously checked even after primers have been designed. This also applies to primer sequences which have been published, for example, the 858L primer in our previous study [2], as well as a reverse primer for exon 9 of the LDLR gene which contains a SNP at the seventh position from the 3′ end [9], and a forward primer for exon 12 of the LDLR gene which contains SNPs at the fifth and eighth position from
40
E. Laios, K. Glynou / Clinical Biochemistry 41 (2008) 38–40
the 3′ end [10]. To our knowledge this is the first report of ADO leading to incorrect genotyping in the LDLR gene.
[4]
Acknowledgments [5]
Clinical diagnosis of familial hypercholesterolemia patients was provided by Dr. Euridiki Drogari, Unit of Metabolic Diseases, Choremio Research Laboratory, University of Athens, 1st Department of Pediatrics, “Aghia Sophia” Children’s Hospital, Athens. We thank Dr. Peristera Paschou from the Democritus University of Thrace for the useful discussions.
[6]
[7]
References [8] [1] Goldstein JL, Hobbs HH, Brown MS. Familial hypercholesterolemia. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic and molecular bases of inherited disease, 8th ed., vol. II. New York: McGraw Hill; 1995. p. 1981–2030. [2] Laios E, Drogari E. Analysis of LDLR mutations in familial hypercholesterolemia patients in Greece by use of the NanoChip microelectronic array technology. Clin Chim Acta 2006;374(1-2):93–9. [3] Allard D, Amsellem S, Abifadel M, et al. Novel mutations of the PCSK9 gene cause variable phenotype of autosomal dominant hypercholesterol-
[9]
[10]
emia. Hum Mutat 2005;26(5):497 (Erratum in: Hum Mutat 2005;26(6): 592). Humphries SE, Whittall RA, Hubbart CS, et al. Genetic causes of familial hypercholesterolemia in patients in the UK: relation to plasma lipid levels and coronary heart disease risk. J Med Genet 2006;43(12):943–9. Ellard S, Bulman MP, Frayling TM, Allen LIS, Dronsfield MJ, Tack CJ, et al. Allelic drop-out in exon 2 of the hepatocyte nuclear factor-1a gene hinders the identification of mutations in three families with maturityonset diabetes of the young. Diabetes 1999;48:921–3. Zajickova K, Krepelova A, Zofkova I. A single nucleotide polymorphism under the reverse primer binding site may lead to BsmI misgenotyping in the vitamin D receptor gene. J Bone Miner Res 2003;18(10): 1754–7. Heinrich M, Müller M, Rand S, Brinkmann B, Hohoff C. Allelic drop-out in the STR system ACTBP2 (SE33) as a result of mutations in the primer binding region. Int J Legal Med 2004;1189:361–3. Ward KJ, Ellard S, Yajnik CS, Frayling TM, Hattersley A, Venigalla PNS. Allelic drop-out may occur with a primer binding site polymorphism for the commonly used RFLP assay for the −1131T N C polymorphism of the Apolipoprotein AV gene. Lipids Health Dis 2006;5:11–6. Hobbs HH, Brown MS, Goldstein JL. Molecular genetics of the LDL receptor gene in familial hypercholesterolaemia. Hum Mutat 1992;1: 445–66. Nissen H, Guldberg P, Hansen AB, Petersen NE, Horder M. Clinically applicable mutation screening in familial hypercholesterolemia. Hum Mutat 1996;8:168–77.
Clinical Biochemistry 41 (2008) 38 – 40
Allelic drop-out in the LDLR gene affects mutation detection in familial hypercholesterolemia Eleftheria Laios a,⁎, Kyriaki Glynou b a
Unit of Metabolic Diseases, Choremio Research Laboratory, University of Athens, 1st Department of Pediatrics, “Aghia Sophia” Children’s Hospital, Athens 11527, Greece b Medicon Hellas S.A., Gerakas 15344, Greece Received 4 May 2007; received in revised form 19 August 2007; accepted 28 September 2007 Available online 11 October 2007
Abstract Objectives: Familial hypercholesterolemia is a monogenic disorder caused by mutations in the LDL receptor (LDLR) gene. We observed allelic drop-out during LDLR genotyping and aimed at redesigning mutation detection. Design and methods: The NanoChip microelectronic array technology and PCR restriction fragment length polymorphism analysis were used. Results: Allele drop-out caused false homozygous diagnoses and was overcome using PCR primers without polymorphisms in the primer binding site. Conclusions: This report presents the importance of allele drop-out in LDLR genotyping. © 2007 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. Keywords: Low-density lipoprotein receptor; Hypercholesterolemia; Mutation; Greece
Introduction
Materials and methods
Familial hypercholesterolemia (FH) is a monogenic autosomal dominant disorder characterized by elevated low-density lipoprotein (LDL) cholesterol, tendon xanthomas, and premature coronary heart disease [1]. FH is caused by mutations in the LDL receptor (LDLR) gene [1]. FH clinical diagnosis is based on family history and elevated LDL cholesterol. Extensive genetic screening for diagnosing FH patients is in effect in several countries. Recently, we genotyped four Greek FH patients as Greece II homozygotes. However, LDL cholesterol levels of these patients indicated a heterozygote phenotype. We hypothesized that allele drop-out had caused false homozygous diagnoses in the aforementioned cases.
Genotyping at nucleotide 858 of LDLR exon 6
⁎ Corresponding author. Fax: +30 210 7709415. E-mail address: [email protected] (E. Laios).
Blood collections were performed with informed oral consent of the participants with a clinical diagnosis of FH. Genomic DNA was isolated by use of the QIAamp DNA Blood Mini Kit (Qiagen). For detection of the Greece II mutation (C to A transition at nucleotide 858 of exon 6), DNA was amplified with the forward primer 858U (5′ biotin-CTCTGCGAGGGACCCAAC 3′) and the reverse primer 858L (5′ TTTCCTGGCTGGGGACAA 3′). The PCR product was used for NanoChip analysis according to [2]. Primer 858L resulted in allele drop-out (due to the underlined nucleotide) and was subsequently replaced by primers 858L-2 or 858L-3 which do not bind to regions containing single nucleotide polymorphisms (SNPs). Primer 858L-2 is located 56 bases downstream of 858L and primer 858L-3 is located 386 bases downstream of 858L. For PCR-RFLP analysis, DNAwas amplified with 858U and 858L-2 (5′ CGCCCAGCATCGCTTCATTTTTT 3′). The PCR product
0009-9120/$ - see front matter © 2007 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved. doi:10.1016/j.clinbiochem.2007.09.017
E. Laios, K. Glynou / Clinical Biochemistry 41 (2008) 38–40
was digested with MnlI (NEB), followed by separation of digestion products by agarose gel electrophoresis. Greece II introduces a MnlI restriction site to alleles with the C858A mutation. To repeat NanoChip analysis of samples previously genotyped as wild-type at position 858, DNAwas amplified with 858U and 858L-3 (5′ CACAGGGTGGGCAGAGTG 3′) followed by mutation detection according to [2]. Genotyping SNP rs13306513 at nucleotide 976 of LDLR intron 6 The SNP rs13306513 (A or G allele) is located at nucleotide 976 in intron 6 of the LDLR gene. We genotyped the SNP both in the general Greek population and in FH patients carrying the Greece II mutation (C858A). For the general population, saliva samples were collected after informed oral consent of 65 unrelated Greeks. Saliva collection and genomic DNA extraction were performed with the Oragene kit (DNAGenotek). Genomic DNA of FH patients was obtained as described in the Genotyping at nucleotide 858 of LDLR exon 6 section. For PCRRFLP analysis, DNA was amplified with 858U and 858L-2 (5′ CGCCCAGCATCGCTTCATTTTTT 3′). The PCR product was digested with MnlI (NEB), followed by separation of digestion products by agarose gel electrophoresis. The A allele of the SNP introduces a MnlI restriction site. Results Using amplification primers 858U and 858L, we identified four Greece II homozygous FH cases based on the NanoChip assay [2]. We confirmed these results by sequencing the same amplicons with 858L. Sequencing revealed a homozygous genotype in all four cases. However, the clinical phenotype of these patients was not consistent with the molecular results. Investigation of the false homozygous diagnosis resulted in identification of a SNP within the primer binding region of the reverse primer (858L). The SNP (rs13306513 at nucleotide 976; A or G allele) was located at the sixth position from the 3′ end of the primer and resulted in drop-out of the A976 allele. The SNP had not been reported at the time of our experimental design. We redesigned the reverse primer and performed PCR-RFLP analysis of the four pseudo-homozygotes. Using a reverse primer without a SNP in the primer binding region (primer 858L-2), the samples were genotyped as Greece II heterozygotes. According to RFLP, in all four cases, the allele with the Greece II mutation (A858) was G at 976, while the normal C858 allele was A at 976. This is consistent with the allele drop-out (ADO) observation in the four pseudo-homozygotes. The four patients in which ADO was observed belonged to three different families. We have genotyped an additional 33 unrelated Greece II heterozygotes at position 976 and have found no A alleles. Using the redesigned reverse primer 858L-3 and NanoChip analysis, we also repeated genotyping of 50 samples which we had previously characterized as normal at position 858 of exon 6 using PCR primer 858L [2]. Based on the repeated NanoChip analysis, the 50 samples appeared as wild type. This indicated that the samples had not been falsely genotyped as normals (ADO due
39
to A976 on an A858 allele could have resulted in a Greece II heterozygote falsely appearing as a normal). The prevalence of the SNP (rs13306513; A or G allele) at position 976 in the general Greek population was estimated by studying a total of 130 alleles from 65 unrelated individuals. We identified 125 G alleles and 5 A alleles. The allele frequency was estimated as 0.96 for the G allele and 0.04 for the A allele. Genotype frequencies were 0.96 for GG, 0.04 for AG, while AA was not found. Linkage disequilibrium (LD) between the two SNPs was estimated in a population consisting of the 36 unrelated heterozygote C858A patients and the 65 unrelated individuals from the general population (D′ = 0.031, r2 = 0). However, it is difficult to conclude that the two SNPs are not co-segregating since the LOD score associated with the LD calculations was very low (LOD = 0). This is probably due to the very low frequency of the rare allele at position 976 and depicts an inherent problem in LD calculation using the metrics of D′ and r2 when allele frequencies at the studied SNPs are skewed. Discussion Our results show that the presence of a polymorphism in the primer binding site leads to allelle drop-out in exon 6 of the LDLR gene, creating a potential problem in FH molecular diagnosis. ADO caused four FH heterozygotes to be falsely genotyped as homozygotes. In general, misgenotyping of an FH patient with a clinical heterozygote phenotype as a homozygote with two mutant alleles is easily detected since this would require a severe clinical phenotype and two heterozygous FH parents. However, misgenotyping of an FH patient with a clinical heterozygote phenotype as a normal with two wild-type alleles would lead to unnecessary analysis of the remaining LDLR gene for mutation identification. In addition to the LDLR gene, the apolipoprotein B100 gene and the proprotein convertase subtilisin/kexin type 9 (PCSK9) gene are known to cause the same autosomal dominant hypercholesterolemic phenotype. In most populations, only about 70–80% of heterozygous FH phenotypes are explained by mutations in the LDLR and apolipoprotein B100 genes, while PCSK9 accounts for a small percentage of the remaining nonLDLR/non-apolipoprotein B100 cases [3,4]. Although it is possible that additional genes are responsible for the autosomal dominant hypercholesterolemia phenotype, it is also likely that diagnoses may be missed due to ADO. Several studies have shown that polymorphisms within the primer binding site can result in ADO due to preferential amplification of one allele [5–8]. Therefore, primer design should include checking more than one SNP database for polymorphisms in the binding site. Since these databases are frequently updated, they should be continuously checked even after primers have been designed. This also applies to primer sequences which have been published, for example, the 858L primer in our previous study [2], as well as a reverse primer for exon 9 of the LDLR gene which contains a SNP at the seventh position from the 3′ end [9], and a forward primer for exon 12 of the LDLR gene which contains SNPs at the fifth and eighth position from
40
E. Laios, K. Glynou / Clinical Biochemistry 41 (2008) 38–40
the 3′ end [10]. To our knowledge this is the first report of ADO leading to incorrect genotyping in the LDLR gene.
[4]
Acknowledgments [5]
Clinical diagnosis of familial hypercholesterolemia patients was provided by Dr. Euridiki Drogari, Unit of Metabolic Diseases, Choremio Research Laboratory, University of Athens, 1st Department of Pediatrics, “Aghia Sophia” Children’s Hospital, Athens. We thank Dr. Peristera Paschou from the Democritus University of Thrace for the useful discussions.
[6]
[7]
References [8] [1] Goldstein JL, Hobbs HH, Brown MS. Familial hypercholesterolemia. In: Scriver CR, Beaudet AL, Sly WS, Valle D, editors. The metabolic and molecular bases of inherited disease, 8th ed., vol. II. New York: McGraw Hill; 1995. p. 1981–2030. [2] Laios E, Drogari E. Analysis of LDLR mutations in familial hypercholesterolemia patients in Greece by use of the NanoChip microelectronic array technology. Clin Chim Acta 2006;374(1-2):93–9. [3] Allard D, Amsellem S, Abifadel M, et al. Novel mutations of the PCSK9 gene cause variable phenotype of autosomal dominant hypercholesterol-
[9]
[10]
emia. Hum Mutat 2005;26(5):497 (Erratum in: Hum Mutat 2005;26(6): 592). Humphries SE, Whittall RA, Hubbart CS, et al. Genetic causes of familial hypercholesterolemia in patients in the UK: relation to plasma lipid levels and coronary heart disease risk. J Med Genet 2006;43(12):943–9. Ellard S, Bulman MP, Frayling TM, Allen LIS, Dronsfield MJ, Tack CJ, et al. Allelic drop-out in exon 2 of the hepatocyte nuclear factor-1a gene hinders the identification of mutations in three families with maturityonset diabetes of the young. Diabetes 1999;48:921–3. Zajickova K, Krepelova A, Zofkova I. A single nucleotide polymorphism under the reverse primer binding site may lead to BsmI misgenotyping in the vitamin D receptor gene. J Bone Miner Res 2003;18(10): 1754–7. Heinrich M, Müller M, Rand S, Brinkmann B, Hohoff C. Allelic drop-out in the STR system ACTBP2 (SE33) as a result of mutations in the primer binding region. Int J Legal Med 2004;1189:361–3. Ward KJ, Ellard S, Yajnik CS, Frayling TM, Hattersley A, Venigalla PNS. Allelic drop-out may occur with a primer binding site polymorphism for the commonly used RFLP assay for the −1131T N C polymorphism of the Apolipoprotein AV gene. Lipids Health Dis 2006;5:11–6. Hobbs HH, Brown MS, Goldstein JL. Molecular genetics of the LDL receptor gene in familial hypercholesterolaemia. Hum Mutat 1992;1: 445–66. Nissen H, Guldberg P, Hansen AB, Petersen NE, Horder M. Clinically applicable mutation screening in familial hypercholesterolemia. Hum Mutat 1996;8:168–77.