| Human Molecular Genetics | Pages |
©1999 Oxford University Press |
Mutation -59c->t in repeat 2 of the LDL receptor promoter: reduction in transcriptional activity and possible allelic interaction in a South African family with familial hypercholesterolaemia
Introduction
Results
Discussion
Materials And Methods
Details of the index patient and study subjects
DNA analysis
Reporter vector constructs and transient transfection assays
Acknowledgements
References
Mutation -59c->t in repeat 2 of the LDL receptor promoter: reduction in transcriptional activity and possible allelic interaction in a South African family with familial hypercholesterolaemia
Received April 27, 1999; Revised and Accepted July 9, 1999
The low-density lipoprotein receptor (LDLR) plays a major role in cholesterol homeostasis. Mutations in the regulatory region of the LDLR gene, although rare, have been shown to alter transcriptional activity of the gene and can cause familial hypercholesterolaemia (FH). In this study, a transition (c->t) was identified at nucleotide position -59 within repeat 2 of the LDLR promoter in a South African FH patient of mixed ancestry. By screening 17 family members of the index case for this promoter mutation, two additional single base changes (-124c->t and -175g->t) were identified, located at recently described cis-acting regulatory sequences of the LDLR promoter. Both the -59c->t and the -124c->t transitions were identified in the normocholesterolaemic son of the index patient. Reporter plasmids containing the normal and mutant promoter fragments were constructed by directional cloning. Transcription studies using a luciferase reporter system demonstrated that the -59c->t mutation significantly reduces promoter activity in both the presence and absence of sterols (~40% of normal activity), while the -124c->t variant increases transcription (~160%) of the LDLR gene. The intra-familial phenotypic variability observed amongst individuals with the -59c->t mutation can probably be ascribed to allelic interaction, suggesting that variation in the LDLR promoter region may contribute sig- nificantly to the phenotypic expression of FH-related mutations in populations where these mutations prevail.
INTRODUCTION
Familial hypercholesterolaemia (FH) is an autosomal dominant disorder affecting the regulation of cholesterol homeostasis. Clinical and biochemical features of FH include xanthomata, premature coronary heart disease (CHD) and elevated plasma cholesterol (1). Most FH-related mutations identified to date are located in the coding region of the low-density lipoprotein receptor (LDLR) gene, while mutations in the promoter region appear to be rare (2-5).
The essential regulatory elements of the LDLR gene lie within ~200 bp upstream of the transcription initiation site, and three imperfect direct repeats are largely responsible for promoter activity (6). Repeats 1 and 3 bind Sp1, a trans-acting transcription factor, which promotes transcription of the LDLR gene in the presence and absence of sterols (6,7). The 10 bp core sequence of repeat 2, designated the sterol regulatory element (SRE-1), is essential for high levels of transcription of the LDLR gene (8-10). In the case of sterol depletion, SRE-1 interacts with the essential transcription binding proteins (SREBP) to induce transcription of the LDLR gene (11-13), whilst responsible for sterol-mediated repression of the gene when cellular cholesterol levels are high (8,9). Mutations in the core of repeat 2 have been shown to reduce transcription significantly only in the absence of sterols (8,9). Two additional cis-acting regulatory elements have recently been identified and are designated footprinting 1 (FP1) and footprinting 2 (FP2) elements (14). FP1 and FP2, spanning nucleotide intervals -145 to -126 and -187 to -175, respectively, are deemed essential for maximal induction of transcription. In 1997, Dhawan et al. (15) suggested that FP1-induced transcription might be through interaction with SRE-1. A variety of substances, such as cytokines, growth factors and hormones, has also been reported to influence regulation of transcription of the LDLR gene (16,17).
In this study, we identified single base changes at nucleotide positions -59(c->t), -124(c->t) and -175(g->t) of the LDLR gene promoter in a South African family of mixed ancestry. Of these, only the -59c->t mutation located in repeat 2 was associated with the FH phenotype. Together with the sequence changes described here, nine different defects/polymorphisms have now been reported in the LDLR promoter region (3,4,10,18,19).
RESULTS
A base change (c->t) at nucleotide position -59 was identified by combined heteroduplex and single-strand conformation polymorphism (HEX-SSCP) analysis (20) in repeat 2 of the LDLR gene promoter in a South African FH patient of Afro-Euro-Malay origin. Subsequent screening for the mutation in 17 additional relatives revealed the presence of mutation -59c->t in the brother (II-5) and two sons (III-8, III-11) of the index case (Fig. 1A). Individuals II-5 and III-11 presented with total cholesterol (TC) levels above the 80th percentile for age and gender (21), while individual III-8 demonstrated an apparently normal lipid profile (Table 1). Interestingly, further mutation analysis in this individual (III-8) revealed an additional single base change (c->t) at nucleotide position -124 (Fig. 1B, lane 6). Furthermore, we detected a base change (g->t) at nucleotide position -175 in the samples of two normo- cholesterolaemic nieces (III-2, III-5) of the index patient. DNA sequence analysis of the three promoter mutations detected in the family is illustrated in Figure 2. The lipid profiles of the mutation-negative daughters of the index patient were within the normal range, while her mutation-negative siblings had moderately raised TC concentrations (below the 70th percentile) for age and gender according to Rossouw et al. (21).
Figure 1. Mutation analysis in the index family. (A) Structure of the family. Family members with the apparently disease-related mutation -59c->t, including the normocholesterolaemic son (III-8) of the index patient (identified by an arrow), are indicated by half-filled symbols. (B) Heteroduplex analysis of the LDLR promoter region. DNA of a control individual with the -124c->t variant was loaded in lane 1. The index patient (lane 2) and one of her sons (lane 3) were heterozygous for mutation -59c->t, while the eldest son (lane 6) was heterozygous for both -59c->t and -124c->t. The promoter mutations were absent in the two daughters (lanes 4 and 5).
Figure 2. DNA sequence analysis in members of the index family with mutations -59c->t, -124c->t and -175g->t. Each base change is depicted in the sequence by an asterisk (*).
Table 1. Characteristics of the index case and several relatives
| Subjects | Gender | Age (years) | TC (mmol/l) | TG (mmol/l) | HDLC (mmol/l) | LDLC (mmol/l) | Mutation | Apo E genotype |
| II-2 | F | 71 | 6.5 | 2.7 | 0.9 | 4.4 | - | 3/3 |
| II-3 | F | 68 | 7.7 | 3.0 | 0.9 | 5.4 | - | 3/4 |
| II-5 | M | 63 | 9.2 | 1.7 | 1.2 | 7.2 | -59c->t | 3/3 |
| II-6 | F | 60 | 5.9 | 1.3 | 1.5 | 3.8 | - | 3/3 |
| II-7 | F | 57 | 7.3 | 1.0 | 1.4 | 5.4 | - | 3/4 |
| II-10a | F | 54 | 9.5 | 0.9 | 1.3 | 7.8 | -59c->t | 3/3 |
| II-11 | M | 60 | 3.0 | 0.9 | 0.9 | 1.7 | - | 3/4 |
| III-2 | F | 44 | 4.0 | 0.8 | 1.1 | 2.5 | -175g->t | 3/3 |
| III-3 | F | 31 | 3.7 | 1.0 | 1.2 | 2.0 | - | 3/3 |
| III-4 | M | 38 | 5.2 | 2.3 | 1.0 | 3.1 | - | 3/3 |
| III-5 | F | 42 | 4.5 | 1.0 | 1.0 | 3.0 | -175g->t | 3/3 |
| III-6 | F | 37 | 6.1 | 1.2 | 0.8 | 4.7 | - | 3/3 |
| III-7 | F | 42 | 7.2 | 0.9 | 0.7 | 6.1 | - | 4/4 |
| III-8 | M | 36 | 4.9 | 0.5 | 2.1 | 2.6 | -59c->t/-124c->t | 3/3 |
| III-9 | F | 30 | 5.3 | 0.5 | 0.8 | 4.3 | - | 2/3 |
| III-10 | F | 23 | 4.5 | 0.5 | 1.3 | 3.0 | - | 3/3 |
| III-11 | M | 15 | 5.1 | 0.7 | 1.6 | 3.2 | -59c->t | 3/3 |
| IV-1 | M | 23 | 4.4 | 0.9 | 1.0 | 3.0 | - | 3/3 |
aIndex case.
DNA screening of an additional 151 FH heterozygotes from the same population for the presence of promoter variants led to the identification of a single patient with the -175g->t base change. Subsequent DNA screening of this individual for mutations in the coding region of the LDLR gene revealed a G->A base change (E237K) in exon 5 (R. Thiart and H. Nissen, unpublished data). In contrast, no disease-causing mutation could be identified in the coding region of the FH index patient with the -59c->t mutation. We also failed to identify an LDLR gene mutation in the hypercholesterolaemic sister (TC > 7 mmol/l and normal triglycerides) of the index case (II-7), who tested negative for mutation -59c->t. This individual was selected for extensive analysis of the LDLR gene (together with the index case) to screen for another genetic factor underlying the hypercholesterolaemia in the family. Mutations -124c->t (previously detected at a low frequency in Africans) and -59c->t were absent in 60 healthy blood donors of mixed ancestry included in this study, while the promoter variant at nucleotide position -175 was detected at a heterozygote frequency of 13% (8/60) in the control population.
In order to determine the possible allelic effects of the sequence changes identified at nucleotide positions -59 and -124 in individual III-8, PCR-amplified products encompassing nucleotide interval -244 to +55 of the wild-type and mutated promoters were cloned into a luciferase reporter vector and transiently transfected into human hepatoma (HepG2) cells. High levels of transcription were observed for the wild-type promoter, while the promoterless vector (pGL3 Basic) showed virtually no effect in the HepG2 cells. In comparison, the -59c->t transition significantly reduced transcription of the LDLR gene promoter to ~40% of normal activity, while the -124c->t base change increased promoter activity to ~160% of normal in sterol depleted cells (P < 0.05). In HepG2 cells supplemented with sterol-containing medium, transcription of the LDLR gene decreased to ~16%, while transcription of the mutant promoters was reduced to 10% (-59c->t) and 50% (-124c->t), respectively (Fig. 3).
Figure 3. Analysis of wild-type (N) and mutated (-59c->t, -124c->t) LDLR promoter activity under transient transfection conditions. Luciferase gene activities were normalized against [beta]-galactosidase activity.
Apolipoprotein (apo) E genotyping was performed in an attempt to determine whether mutation -59c->t exhibits only a mild effect on LDLR function that is exacerbated by the E4 allele of this polymorphism (22). All individuals with mutations in the LDLR promoter region were found to be homozygous for the neutral E3 allele of the apo E polymorphism (Table 1). Notably, the presence of the cholesterol-raising E4 allele was detected in all three family members (II-3, II-7, III-7) with TC levels raised 7 mmol/l in the absence of mutation -59c->t.
DISCUSSION
The base change (c->t) identified at nucleotide position -59 in the index family represents the first report of a naturally occurring mutation in the 10 bp core sequence of repeat 2 in the LDLR promoter. In vitro results demonstrated that this mutation dramatically diminishes transcription of the LDLR gene in both the absence and presence of sterols. The transfection assays were performed in HepG2 cells, where a reduction of transcriptional activity to ~40% of normal in sterol-deficient cells was observed. It has been shown previously that the nucleotide interval -65 to -56, the core sequence of repeat 2, is essential for high levels of transcription, as well as sterol-mediated repression of the gene (8,9). Interestingly, transversions (c->g/a) incorporated at nucleotide position -59 (8,9) virtually abolished (to ~10% of normal activity) the induction of transcription in the absence of sterols, but did not reduce transcription of the LDLR gene in the presence of sterols as observed with the -59c->t mutation identified in the South African patient. This observation may be explained by the fact that the naturally occurring mutation represents a transition, reflecting the possible significance of the specific base content at a given position within repeat 2. Since a mutation at nucleotide position -59 results in non-binding of nuclear proteins (9), and hypermethylation of guanine -59 is positively associated with activation of LDLR gene transcription in skin fibroblasts (23), we conclude that the -59c->t mutation is the most likely cause of the FH phenotype in the index patient. Further evidence in favour of a causative role of this mutation includes the failure to identify an additional disease-causing mutation in the coding region of the LDLR gene in the index patient, its apparent absence in healthy control individuals and the fact that this site has remained conserved throughout evolution. The -53c->a mutation reported previously in repeat 2 (4) (I.N.M. Day, personal communication) is located at a position unlikely to have a major effect on sterol-mediated regulation of the LDLR gene (8,9).
The -59c->t transition was detected in three family members of the index case. In keeping with the above-mentioned data, her brother (II-5) and youngest son (III-11) with (only) mutation -59c->t, presented with TC levels above the 80th percentile for age and gender. In a follow-up study of lipid determinations (data not shown) performed after a 10 month period in nine (II-5, II-7, II-10, II-11, III-2, III-4, III-8, III-11, IV-1) relatives, TC concentration in the 15-year-old son (III-11) of the index patient with mutation -59c->t was only moderately raised (4.6 mmol/l; >60th percentile), which can probably be ascribed to altered expression of FH-related mutations during childhood (24). The follow-up TC determinations also confirmed the FH status of the index case's brother (II-5; TC 8.4 mmol/l), the non-FH status of her sister (II-7; 6.4 mmol/l), and the normal lipid profile of her eldest son (III-8; 4.9 mmol/l) carrying both mutations -59c->t and -124c->t. Plasma cholesterol levels remained normal in subjects II-11, III-2, III-4 and IV-1 without mutation -59c->t, which is consistent with the finding that this mutation co-segregates with the FH phenotype in family members without the sequence variant at nucleotide position -124(c->t).
Although it was not possible to study the influence of the numerous environmental (e.g. diet) and genetic factors that may contribute to the abnormal lipid profile in the index family, possible allelic effects imposed by the common apo E polymorphism were excluded in those subjects with mutations in the LDLR promoter region. The presence of the apo E4 allele, previously shown to be associated with raised plasma cholesterol concentrations (22), may nevertheless explain the raised TC levels in mutation-negative hypercholesterolaemics II-3, -7, III-7 (TC > 7 mmol/l).
The normal plasma cholesterol levels detected in the mutation-positive son (III-8) of the index patient can probably be ascribed to allelic interaction between the mutations at nucleotide positions -59 and -124, since the transfection results revealed a statistically significant increase in promoter activity for the -124c->t construct alone, to ~160% of normal in sterol depleted cells. The -124c->t variant is located adjacent to the FP1 cis-acting regulatory element (position -126 to -144), previously implicated in maximal induction of the human LDLR gene transcription in response to low cellular cholesterol levels. Demonstration in this study that the variant (c->t) at nucleotide position -124 increases LDLR transcriptional activity indicates that the boundary of the FP1 enhancer sequence should probably be extended to include nucleotide position -124. It is noteworthy that this variant was absent in the FH patients analysed. This observation may be due to chance, but as illustrated in the index family, it is highly unlikely that an individual with this apparently favourable variant as well as a disease-causing LDLR gene mutation would present with elevated plasma cholesterol levels; except maybe when the two sequence changes occur on the same chromosome. Interestingly, DNA screening of >1000 individuals from eight different ethnic groups demonstrated the presence of variant -124c->t at a low frequency (1-3%) in populations with an African genetic element, while apparently absent in Caucasians (25).
Identification of variant -175g->t in normocholesterolaemic individuals, as well as in a single proband (who died recently at the age of 50 years of a heart attack) with an FH-related mutation in the coding region of the LDLR gene, indicates that this base change does not cause the FH phenotype. However, as suggested by our preliminary data obtained in the South African Black population (26), it is possible that the presence of the -175g->t variant imposes susceptibility or an increased risk for the development of symptomatic FH in patients with other disease-related mutations. In a study of the coding region of the LDLR gene in South African FH patients of mixed ancestry, we demonstrated recently that Caucasian admixture has contributed significantly to the disease phenotype in this indigenous population (27). Since the -175g->t polymorphism has not been detected in Caucasians, we postulate that the significantly lower frequency (1/151) of this variant in FH patients of mixed ancestry compared with controls from the same population (~13%), is probably a reflection of the genetic profile at the LDLR gene locus in FH patients as a consequence of admixture linkage disequilibrium. Interestingly, variant -175g->t is located within a recently defined FP2 cis-acting regulatory element, and disrupts a putative binding site for the multifunctional transcription factor F-ACT1 (YY1) (14). However, whether the presence of the rare -175t allele increases CHD risk in FH patients of mixed ancestry was not the focus of this study, and therefore further data are not included in this report.
This study highlights the role of the SREBP in the regulation of LDLR gene transcription and suggests that the -59c->t mutation, leading to reduced transcriptional activity both in the presence and absence of sterols, is the causative FH mutation in the index family. In contrast to clinical FH homozygotes who present with severely elevated plasma cholesterol levels due to the presence of two mutant LDLR alleles, we have identified an individual with two point mutations in the promoter region of the gene, whose cholesterol concentration is within the normal range. Further studies are warranted to determine whether the normocholesterolaemic status in this `compound heterozygote' may provide the sought-after in vivo evidence for interaction between FP1 and SRE-1, as postulated previously by Dhawan et al. (15).
MATERIALS AND METHODS
Details of the index patient and study subjects
The index case is a 54-year-old South African woman of mixed ancestry (San, Khoi, African Negro, Madagascar, Javanese and Western European origin), clinically diagnosed with FH. She presented at age 44 with angina, xanthomata and arcus corneae, smoked ten cigarettes per day and had no documented family history of premature CHD. At this age, her pre-treatment TC concentration was 7.3 mmol/l, TG 0.6 mmol/l, high-density lipoprotein cholesterol (HDLC) 1.6 mmol/l and low-density lipoprotein cholesterol (LDLC) 5.5 mmol/l. For inclusion in this study, her plasma lipid levels were measured again together with that of 17 additional family members, using standard techniques (Table 1).
Follow-up mutation screening was performed in 151 unrelated FH heterozygotes of mixed ancestry attending the Groote Schuur Hospital lipid clinic. The selection of the samples was based on previously described criteria for a diagnosis of FH (24), including TC > 7 mmol/l, the presence of tendon xanthomas and/or premature CHD in the index patient or first-degree relatives. DNA of 60 healthy blood donors from the same population was also included. All blood samples were obtained with informed consent and the study protocol was approved by the appropriate Institutional Ethics Review Committee.
DNA analysis
Genomic DNA extracted from whole blood was amplified by PCR on an OmniGene Thermal Cycler (Hybaid, UK) using LDLR promoter primers 5[prime]-GAGGCAGAGAGGACAATGGC-3[prime] (forward) and 5[prime]-CCACGTCATTTACAGCATTTCAATG-3[prime] (reverse). PCR products were denatured and fragments resolved on a low cross-linked polyacrylamide gel for combined HEX-SSCP analysis (20) of the LDLR gene. Fragments demonstrating altered mobility were sequenced manually and variation confirmed on an automated sequencer, ABI373 (Perkin Elmer, Foster City, CA). Determination of apo E genotypes was performed using oligonucleotide primers F4 and F6 (28), restriction enzyme digestion with HhaI and gel electrophoresis (29).
Reporter vector constructs and transient transfection assays
Construction of reporter plasmids containing the normal and mutant promoter fragments and transient transfection assays were performed as described previously (18). The calcium phosphate method was used to transfect HepG2 cells with the plasmid DNA. Triplicate wells were assayed for each transfection condition and at least three independent transfection assays were performed for each construct. Luciferase activity was normalized against [beta]-galactosidase activity to correct for transfection efficiency. The in vitro results obtained with the -59t and -124t transcripts described in this study have been verified by Peeters (30) in CHO cells using Transfectam (Promega, Madison, WI) as the transfection reagent.
ACKNOWLEDGEMENTS
Prof. D.R. van der Westhuyzen is thanked for helpful discussion, Sister R. Jooste for family data and S. Jones for technical assistance. This study was supported by the Universities of Stellenbosch and Cape Town, as well as the Medical Research Council. The UNESCO foundation and the Federation for Research and Development (FRD) are acknowledged for travel grants, and the Claude Harris Leon Foundation for a doctoral fellowship awarded to C.L.S.
REFERENCES
+These authors contributed equally to this work
§To whom correspondence should be addressed. Tel: +27 21 9389441; Fax: +27 21 9317810; Email: mjk{at}gerga.sun.ac.za
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: jnl.info{at}oup.co.uk
Last modification:
Copyright© Oxford University Press, 1999.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. Mozas, R. Galetto, M. Albajar, E. Ros, M. Pocovi, and J. C. Rodriguez-Rey A mutation (-49C>T) in the promoter of the low density lipoprotein receptor gene associated with familial hypercholesterolemia J. Lipid Res., January 1, 2002; 43(1): 13 - 18. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||