|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Lipid Research, Vol. 44, 1850-1858, October 2003 Identification and characterization of LDL receptor gene mutations in hyperlipidemic Chinese
* Department of Biology, National Taiwan Normal University, Taipei, Taiwan Published, JLR Papers in Press, July 1, 2003. DOI 10.1194/jlr.M200470-JLR200
1 To whom correspondence should be addressed. e-mail: t43019{at}cc.ntnu.edu.tw
DNA screening for LDL receptor mutations was performed in 170 unrelated hyperlipidemic Chinese patients and two clinically diagnosed familial hypercholesterolemia patients. Two deletions (Del e3-5 and Del e6-8), eight point mutations (W-18X, D69N, R94H, E207K, C308Y, I402T, A410T, and A696G), and two polymorphisms (A370T and I602V) were identified. Of these mutations, C308Y and Del e6-8 were found in homozygosity, and D69N and C308Y were seen in unrelated patients. The effects of mutations on LDL receptor function were characterized in COS-7 cells. The LDL receptor level and activity were close to those of wild type in A696G transfected cells. A novel intermediate protein and reduction of LDL receptor activity were seen in D69N transfected cells. For R94H, E207K, C308Y, I402T, and A410T mutations, only 2064% of normal receptor activities were seen. Conversely, Del e3-5 and Del e6-8 lead to defective proteins with 013% activity. Most of the mutant receptors were localized intracellularly, with a staining pattern resembling that of the endoplasmic reticulum and Golgi apparatus (D69N, R94H, E207K, C308Y, and I402T) or endosome/lysosome (A410T and Del e6-8). Molecular analysis of the LDL receptor gene will clearly identify the cause of the patient's hyperlipidemia and allow appropriate early treatment as well as antenatal and family studies.
Abbreviations: CAD, coronary artery disease; FH, familial hypercholesterolemia; SSCP, single-strand conformation polymorphism Supplementary key words low density lipoprotein receptor mutation cDNA expression haplotype analysis
The concentration of plasma cholesterol is regulated mainly by the LDL receptor pathway, in which circulating LDL is taken into the cells by receptor-mediated endocytosis (1). A high LDL level due to defect in the LDL receptor is responsible for familial hypercholesterolemia (FH), an autosomal dominant disorder with estimated frequencies of one in 500 in most populations (2). In addition to a rise in the concentration of LDL cholesterol in blood, FH is frequently associated with tendon xanthomata and an increased risk of coronary artery disease (CAD). Due to a great biochemical and clinical variability among heterozygote carriers, measurements of plasma total cholesterol and LDL cholesterol do not allow a clear detection of individuals with FH (2). The homozygous FH occurs in about one in 1 million individuals and is more severely affected than the heterozygote carrier. The LDL receptor gene consists of 18 exons spanning 45 kb on chromosome 19p13 (3). The 5.3 kb mRNA encodes a mature protein of 839 amino acids (4). Overall, more than 800 mutations, including gross deletions, minor deletions, insertions, point mutations, and splice-site mutations, scattered over the LDL receptor gene have been reported (5). Among these, deletions or insertions at the LDL receptor gene involve Alu sequences (6, 7). LDL receptor mutations can be studied by Western, flow cytometry, and fluorescence microscopy analyses of mutated cDNA-transfected COS cells (810). These mutations affect the synthesis (class 1), posttranslational processing (class 2), ligand binding activity (class 3), internalization (class 4), or recycling (class 5) of the LDL receptor. Defects of classes 2 and 4 can be either complete or partial. Heterogeneous mutations seem to have caused FH most in individual families (11), but in certain populations, a few mutations still prevail (1214). Mutations in the LDL receptor gene are known to exist in Chinese FH patients in China, Hong Kong, Canada, and Southeast Asia (1519). There is no definite common mutation in Chinese populations due to a founder effect. In this study, 170 unrelated hyperlipidemic Chinese and two clinically diagnosed FH patients were screened by multiplex polymerase chain reaction (PCR), long PCR, and single-strand conformation polymorphism (SSCP) analyses to detect mutations in the LDL receptor gene. Two gross deletions, six point mutations, and two polymorphisms were found. Additionally, the haplotype at nine polymorphic sites and the effect of mutations on LDL receptor function were examined.
Subjects and lipids Blood samples were collected after an overnight fast from 170 unrelated hyperlipidemic patients, ages 3174 years (mean, 58 years), attending the lipid clinic of the Veterans General Hospital, Taipei and Wei Gong Memorial Hospital. Of these subjects, 112 (70 men and 42 women) had primary type IIa hyperlipidemia, and 58 (40 men and 18 women) had type IIb hyperlipidemia. These patients had cholesterol concentrations >6.20 mmol/l (range, 6.2111.32 mmol/l) without tendon xanthomas. Type IIa patients had triglyceride values <2.27 mmol/l, and type IIb patients had concentrations above this value. Patients with familial defective apolipoprotein B (apoB) were excluded by testing for the apoB-3500 mutation (20, 21). In addition, a 12-year-old girl and a 40-year-old woman, who were clinically diagnosed with homozygous FH based on markedly increased concentrations of plasma LDL cholesterol (16.61 and 17.16 mmol/l, respectively) and the presence of skin and tendon xanthomas, were included in this study. Family members of patients with identified LDL receptor mutations and 100 unrelated Chinese subjects with normal fasting plasma lipid levels (cholesterol concentrations <5.17 mmol/l) were also recruited for the study with their consent. The procedures followed were in accordance with the current revision of the Helsinki Declaration of 1975. Blood samples from fasting subjects were drawn for DNA extraction and lipid measurements as described (20, 21).
Mutation analysis
PCR products (3 µl) were mixed with an equal volume of 95% formamide buffer for SSCP analysis (21). Fragments showing mobility shifts of single strands were sequenced directly using the ABI Prism Dye Terminator Kit and an ABI PrismTM 377 DNA Sequencer (Applied Biosystems). All point mutations were verified by restriction analysis (Table 2) of PCR amplified products.
If the presence of a large deletion was indicated by electrophoresis of multiplex PCR products, a long PCR amplification was then performed (Table 1, PCR H and L). A 50 µl reaction contained 500 ng DNA, 0.4 µmol/l of each primer, 0.2 mmol/l deoxynucleotide triphosphates, 1.0 mmol/l MgCl2, and 1 unit of ELONGASE enzyme mix (Gibco Brl). Amplification was set as denaturation at 94°C for 1 min, and primer annealing and extension at 68°C for 10 min. Shorter (deleted) fragments were subcloned into pGEM-T Easy (Promega) and sequenced as described.
cDNA constructs
Expression studies For flow cytometric analysis, cells were harvested by trypsin treatment 48 h after transfection, and cells in suspension were incubated with 0.4% heparin for 30 min to remove any surface-bound LDL. To quantitate the amount of receptor on the cell surface, the cells were incubated for 30 min at 4°C in primary antibody diluted 1:60 (v/v) in PBS, washed in PBS, and incubated further for 30 min at 4°C in FITC-conjugated goat anti-rabbit antibody (Zymed) also diluted 1:60 (v/v) in PBS. To measure the LDL receptor function, the cells were incubated for 30 min at 4°C in DiI-LDL (Molecular Probes) diluted in DMEM to a concentration of 4 µg/ml, washed with PBS, and incubated further for 30 min at 37°C. All cells were subsequently washed and resuspended in PBS for fluorescence activated cell sorter analysis. Cells were analyzed in a FACStar flow cytometer (Becton-Dickinson) equipped with an argon laser operating at 488 nm (antibody) or 514 nm (DiI). A forward scatter gate was established to exclude dead cells and cell debris from the analysis. 104 cells were analyzed in each sample. The intracellular receptor protein was visualized by staining cells fixed in 4% paraformaldehyde and permeabilized in 70% ethanol. Cells were incubated for 60 min at room temperature with primary antibody diluted 1:50 in 1% BSA in TBS, washed in PBS, and incubated for 60 min at room temperature in FITC-conjugated secondary antibody diluted 1:50 in 1% BSA in TBS, washed in PBS. Lysosomes were detected using lysotracker (Molecular Probes) at 0.1 µmol/l for 60 min. The stained cells were analyzed using a Leica TCS confocal laser scanning microscope.
Haplotype analysis
Identification of point mutations The promoter region and 18 exons of the LDL receptor gene were amplified and screened for mutations by SSCP and DNA sequencing. Six reported (5) polymorphic changes, intron 7 SmaI, exon 8 StuI, exon 11 AciI, exon 12 HincII, exon 13 AvaII, and exon 15 MspI, were detected (data not shown). Additionally, seven reported (5) point mutations (W-18X, D69N, R94H, E207K, C308Y, I402T, and A410T) and two novel variations (I602V and A696G) were identified (Table 2). The I602V variant was seen in two normolipidemic controls in addition to three hyperlipidemic patients. This then can be presumed to be a polymorphism (1.0% allele frequency). Other point changes, however, were not detected in our normolipidemic controls.
Identification of deletion mutations
Expression of LDL receptor cDNA mutants The effects of mutations on LDL receptor function were investigated by transient expression in COS-7 cells. Approximately equal levels of LDL receptor mRNA derived from each variant was expressed as quantitated by RT-PCR (data not shown). The LDL receptor protein that derived from each variant was examined by immunostaining of Western blot with LDL receptor polyclonal antibody (Fig. 3) . No nonspecific polypeptide was detected in pcDNA3 vector-transfected cells (Fig. 3, lane 1). One hundred sixty kilodalton mature and 120 kDa precursor proteins, as well as an intermediate form, were detected in wild-type receptor cDNA-transfected cells (Fig. 3, lane 2). A novel intermediate protein that is probably a degradation product and an apparent reduction in mature receptor protein were seen in D69N transfected cells (Fig. 3, lane 3). Although the same molecular weight receptor proteins were seen with R94H, E207K, C308Y, I402T, and A410T mutations, the amounts of mature protein detected were reduced (Fig. 3, lanes 48). For A696G, both the processing and the level of the mature protein were close to those of wild type (Fig. 3, lane 9). Conversely, no receptor protein for Del e3-5 and a defect receptor protein for Del e6-8 were detected (Fig. 3, lanes 10 and 11).
Analyses of antibody-labeled cells and activity measurements revealed that A696G encodes receptors in amounts and activity comparable with the wild-type receptor (98% and 93%, respectively). Del e3-5 encodes receptors that cannot be detected and have no detectable residual activity. All other mutations impair the function, encoding receptors in reduced amounts ( 13%64%) and displaying 12%64% residual activity (Table 3).
Fluorescence microscopy of cells expressing the wild-type or the mutant LDL receptor was performed. As shown in Fig. 4A , the wild-type and A696G receptor proteins were located in a ring-shaped structure representing the cell surface, whereas most of the mutant receptors accumulated intracellularly. Colocalization studies using a lysosome staining dye (Fig. 4B) further established that A410T and Del e6-8 mutant proteins are retained in the endosomal/lysosomal regions, whereas D69N, R94H, E207K, C308Y, and I402T produce proteins that seem to be localized in the endoplasmic reticulum (ER) (Fig. 4C).
Lipoprotein concentrations and family studies The clinical and biochemical features of the nine FH heterozygotes and two FH homozygotes are shown in Table 4. The plasma total cholesterol levels of these heterozygotes ranged from 6.21 to 10.03 mmol/l. Four of them (D102, A133, L41, and A144) had evidence of atherosclerotic disease. It was observed that both FH homozygotes had tendon xanthomas and atherosclerosis, in addition to high values of cholesterol (16.61 and 17.16 mmol/l). Currently, D254 is being treated by periodical plasmapheresis.
A total of 33 family members of patients A133, D63, D102, and D254 gave consent to be screened for the identified D69N, C308Y, W-18X, or Del e6-8 mutation. The D69N mutation was found in A133's son and grandson, although their cholesterol values were lower than 6.20 mmol/l (Fig. 5A) . When D63's family was recruited and screened, results showed that both of her parents and three of her siblings also carry the mutation, which was cosegregated with hypercholesterolemia in this family (Fig. 5B). None of the daughters inherited the W-18X mutation from patient D102 (Fig. 5C). Patient D254 inherited Del e6-8 from her parents. Among her siblings, four heterozygous and one homozygous deletion were found. Eight of the ten individuals in the third generation also carried the deletion. With the exception of one individual (cholesterol value 4.94 mmol/l), the deletion was cosegregated with hypercholesterolemia in this family (Fig. 5D).
All of the above mutation carriers are informed and followed up at our out-patient clinics. Those carriers with normal phenotype will have periodic biochemistry examination. Lipid-lowering drugs, including fibric acid derivatives and statins, are prescribed for those carriers with elevated values of cholesterol.
Haplotype analysis on the mutant allele
In the 170 unrelated hyperlipidemic patients and two clinically diagnosed FH homozygotes, eight point mutations and two deletions were found in a total of 12 patients. Compared with previous reports (1518), the frequency of mutations identified in hyperlipidemic Chinese is less than expected. The discrepancy of the results may be due to the screening method, population stratification, and polygenic nature of the disease. The SSCP screening method does not detect every possible mutation. The subjects in the studies of Sun et al. (15), Mak et al. (16), Pimstone et al. (17), and Khoo et al. (18) were clinically diagnosed FH patients, whereas the subjects in our study only show elevated cholesterol values. Mutations in other genes, such as LDL receptor adaptor protein and adenosine triphosphate binding cassette transporters, may also contribute to the elevated cholesterol (23). Although the two FH homozygotes were affected similarly to those elsewhere, the heterozygotes were much less severely affected than their counterparts in Western populations, and rarely exhibited tendon xanthomas or premature CAD. The lack of clinical presence has also been observed in Chinese FH heterozygotes in China (15, 19, 24). However, in the Chinese patients in Hong Kong, Canada, and Southeast Asia, heterozygous FH appears to manifest in a way similar to that seen in Western countries (1618). This may be due to traditional low dietary fat consumption in Chinese in Taiwan and China. This indicates that conventional methods for diagnosis of FH, based mainly on lipid determinations and a family history of CAD, may not always allow an accurate diagnosis of the disease in all Chinese. A low-fat diet modulating the phenotype of Chinese patients in Taiwan with heterozygous familial defective apoB-100 has also been reported (20, 21). The W-18X mutation in the signal peptide predicts no LDL receptor protein synthesis (class 1). This mutation has also been seen in compound heterozygous FH patients from China (11, 15). The lack of haplotype reported precludes the prediction of the mutation origin. The D69N mutation affects the highly conserved aspartate residue 69 in the second cysteine-rich repeats in the binding domain (3, 4). The novel intermediate and reduced mature receptor protein (Fig. 3) indicate abnormal processing and delayed transport of newly synthesized receptors to the cell surface (class 2B) similar to D69G and D69Y mutations reported earlier (11, 26). The LDL receptor was retained in the ER with 55% activity (Table 3). The mutation was found in two Chinese patients in Hong Kong and Malaysia, and previously in a British patient (16, 18, 25). Two patients (A133 and L41) originating from Fukien province of China were noted to have this mutation. Because Chinese in Hong Kong and Malaysia originated primarily from the southern Chinese population, the D69N mutation found in Chinese patients in Hong Kong, Malaysia, and Taiwan may have been inherited from a common ancestor. The results of the haplotype analysis also suggest that the D69N alleles of our two patients are likely identical by descent (Table 5). When the family members of A133 were screened for the D69N mutation, two additional carriers were found. However, their cholesterol values were relatively low (4.86 and 5.74 mmol/l, respectively; Fig. 5A) as compared with those Chinese patients in the studies of Mak et al. (16) and Khoo et al. (18) (11.0 and 8.7 mmol/l, respectively) and our index cases A133 (8.04 mmol/l) and L41 (8.22 mmol/l) (Table 4), suggesting an environmental or genetic effect on the phenotypic expression. Both the patient and carrier relatives are kept fully informed and adhere to a regular follow-up scheme. The R94H mutation is located in the third cysteine-rich repeat in the binding domain. Although evolutionarily conserved among human, rat, rabbit, and cow (4, 2729), arginine94 is not conserved among the seven cysteine-rich repeats (3). The mutation results in slower processing and delayed transport (class 2B). The mutation was also found in two Chinese patients and a Japanese patient (18, 30). No prediction of the mutation origin can be made due to the lack of haplotype reported. The E207K mutation is located in the fifth cysteine-rich repeats in the binding domain. The mutation affects the highly conserved negatively charged Ser-Asp-Glu triplet at the COOH-terminus of all ligand binding repeats (3) and results in severely slow processing and delayed transport (class 2B). This mutation occurs at a CpG hot spot in which the G is mutable to A, C, or T (11). Such mutation has been detected in different ethnic groups (11, 12, 16, 17, 31) and might have occurred independently during evolution. For C308Y, the cysteine residue 308 in the cysteine-rich repeat A of the epidermal growth factor (EGF) precursor homology domain is highly conserved and is important for the correct folding of the LDL receptor protein (3). The disruption of the disulfide linkage formation may lead to the reduced mature protein (class 2B) (Fig. 3), similar to other transport defect mutations nearby, C297Y (11) and C317S (13). The LDL receptor activity was reduced to 31% (Table 3). Previously, the C308Y mutation was found in two Chinese patients in Hong Kong (16) and two Chinese patients in Malaysia (18). In patient D63, the same haplotype of two C308Y alleles resulted from consanguineous marriage. This haplotype was the same as that reported [SfaNI+/AvaII-/NcoI-/(dTA)n 10 repeats] in two Chinese patients in Hong Kong (32), but could not be deduced from 95-035 (Table 5). The data suggest limited multiple recurrent origins for C308Y in Chinese populations. Mutations I402T and A410T were between cysteine-rich repeats B and C of the EGF precursor homology domain. The domain serves to position the ligand-binding domain to bind LDL on the cell surface. It is also required for the acid-dependent dissociation of lipoproteins from the receptor in the endosome during receptor recycling. The amount of receptor protein produced by both mutations was reduced (Fig. 3). I402T was reported as pathogenic in mediating uptake and degradation of LDL (33). The LDL receptor was retained in the ER with 54% activity (class 2B) (Table 3). A410T may be associated with a recycling-deficient phenotype (34). The mutant receptor retained 20% activity (Table 3) and gave rise to an endosomal/lysosomal staining pattern (class 5) (Fig. 4). Both I402T and A410T mutations have been reported previously in different ethnic groups (18, 34, 35). Due to the geographical distance and historic ethnic differences, it is probable that the two mutations have arisen independently in Chinese. The A696G in the O-linked sugars domain was novel. The point change was not detected in our normolipidemic controls. A696G is a conservative amino acid substitution, and alanine at 696 is not conserved across species among human, rat, rabbit, and cow (4, 2729). Although patient D162 had an elevated cholesterol value (6.77 mmol/l) in addition to hypertension, A696G is a rare sequence variation that does not affect LDL receptor function (Fig. 3, Table 3). Two amino acid sequence variants in the EGF precursor homology domain were noted. The A370T variant (StuI RFLP) was first reported in the South African Afrikaner population, where the frequency was found to be 8% (36). Among several Chinese groups studied, the A370T was only found once in our patients (0.3% allele frequency). The novel I602V variant (EcoRV RFLP) is a conservative amino acid substitution. Isoleucine602 is a valine at this position of rat, rabbit, and cow LDL receptor proteins (2729). The I602V variant was found with a frequency of 1% in both normal and hyperlipidemic Chinese. Two novel deletions involving Alu sequences (37) were identified and characterized. Del e3-5 is caused by misalignment of Alu sequences in intron 2 and intron 5 and predicted to cause deletion of 209 aa (44252), including cysteine-rich repeats II to VI in the binding domain (3). The resultant defect receptor protein cannot to be detected and is unable to bind with the LDL (class 2A) (Fig. 3, Table 3). A similar misalignment of Alu sequences in introns 5 and 8 is responsible for Del e6-8. The mutation is predicted to cause deletion of 123 aa (252374), including the seventh disulfide-rich repeat in the binding domain and cysteine-rich repeats A and B in the EGF precursor homology domain (3). The resultant defect receptor retained 12% of its activity and was localized in the endosomal/lysosomal regions (class 5) (Table 3, Fig. 4). The same haplotype of two Del e6-8 alleles in D254 resulted from a known consanguineous marriage. The LDL receptor gene contains many highly redundant Alu sequences in its introns and flanking sequences, as well as in the 3' noncoding region (4). A number of characterized deletions or insertions at the LDL receptor gene involve Alu sequences (5). Our results further support the hypothesis that Alu sequences are hot spots for recombinational events at the LDL receptor locus. In summary, among more than 30 mutations noted in the Chinese, recurrent point mutations in unrelated patients were identified (11, 1518, 32, and the present study). Although the incidence rate is too low to implicate common mutations due to a founder effect, the rapid RFLP detection methods may provide a practical direction and approach for further studies of hyperlipidemic Chinese patients. In FH, a complex genetic disease, to clearly define individual gene defects in patients is of prognostic value, making early detection, optimal treatment approaches, and a reduction in the risk factors associated with CAD highly possible.
The authors thank the FH families, without whose participation this study would not have been possible. The authors are grateful to Dr. T. G. Jensen for providing the green fluorescent protein-tagged LDL receptor used to workout fluorescence microscopy procedures. The authors also thank J. E. Curran and A. Vendouris for their assistance in editing this paper. A portion of this work was supported by grant NSC89-2311-B-003-025 from the National Science Council, Executive Yuan, Republic of China. Manuscript received December 16, 2002 and in revised form June 2, 2003.
This article has been cited by other articles:
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||