| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Methods |
Robarts Research Institute, London, Ontario, Canada N6A 5K8
Published, JLR Papers in Press, December 1, 2004. DOI 10.1194/jlr.D400030-JLR200
1 To whom correspondence should be addressed. e-mail: hegele{at}robarts.ca
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
These findings indicate that MLPA may be a useful new adjunctive tool for the molecular diagnosis of FH.
Supplementary key words atherosclerosis • genetics • monogenic disorders • DNA analysis • low density lipoprotein receptor gene
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
1 in 500 people and typically results from a mutation in the LDL receptor gene (LDLR; MIM 606945) (1). Affected individuals have increased plasma LDL cholesterol, which without adequate diagnosis and intervention can increase the risk of fatal coronary heart disease by up to 100-fold compared with the general population (2). Fortunately, treatment with statin drugs can substantially reduce this risk and improve clinical outcome (2), stressing the importance of early and accurate diagnosis.
At the molecular level, FH is now commonly diagnosed using exon-by-exon screening methods, such as exon-by-exon sequence analysis (EBESA) of LDLR from genomic DNA (gDNA) (3, 4). However, this method finds mutations in only 50% of clinically diagnosed FH patents (3, 4). Part of the diagnostic gap is attributable to the heterogeneity of AD FH (5). For instance, HCHOLAD2 (MIM 144010), which results from a missense mutation in APOB affecting the LDL receptor binding domain of apolipoprotein B-100 (apoB-100; MIM 107730) (6, 7), accounts for 5–10% of the AD FH phenotype. A rare FH subtype called HCHOLAD3 (MIM 603776) results from mutations in PCSK9 (MIM 607786), encoding neural apoptosis-regulated convertase-1 (5, 8, 9). A similarly rare autosomal recessive FH subtype called HCHOLAR1 (MIM 603813) results from mutations in ARH (MIM 605747), encoding a putative adaptor for the LDL receptor (10). However, even after accounting for genetic heterogeneity, many clinically diagnosed FH patients have no LDLR mutation with EBESA.
Another possible explanation for the gap in FH molecular diagnosis is inadequate sensitivity of EBESA to detect certain mutation types, such as large deletions or insertions. The larger gDNA alterations create effective hemizygosity for single exons and result in an EBESA profile that is indistinguishable from homozygosity for two normal LDLR alleles. Multiplex ligation-dependent probe amplification (MLPA) is a new analytical method (11) that detects larger gDNA deletions or insertions that would otherwise be overlooked by EBESA (12). We hypothesized that some FH patients with no LDLR mutation detectable by EBESA would have an abnormality detectable using MLPA. Of 70 unrelated FH patients, 44 had LDLR mutations detected by EBESA, including missense, RNA splicing, nonsense, or small deletion mutations. Five had the APOB R3500Q mutation. Among the remaining 21 AD FH patients with no apparent LDLR mutation, we found abnormal LDLR MLPA patterns in 12 and confirmed the deleted sequences in 5. MLPA may be a useful new adjunctive tool for the molecular diagnosis of FH.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
DNA sequence analysis
From gDNA extracted with the Puregene kit (Gentra Systems, Minneapolis, MN), EBESA of the LDLR was performed as described (4) by sequencing both strands of the promoter region and of the translated parts of all 18 exons (primer information and sequences available upon request). In addition, APOB exon 26, which harbored the receptor binding domain of apo B-100, including the recurrent R3500Q mutation in HCHOLAD2, was examined in all patients using direct sequence analysis. Standard DNA sequencing reactions using the Big Dye Terminator cycle sequencing kit version 3.1 (PE-Applied Biosystems, Mississauga, Ontario, Canada) were analyzed using a 3730 Automated DNA Sequencer (PE-Applied Biosystems) according to the protocols of the London Regional Genomics Centre (www.lrgc.ca).
MLPA reaction
The SALSA P062 LDLR MLPA kit was obtained from MRC-Holland (Amsterdam, the Netherlands). The principles and stages of MLPA have been previously described (11). Briefly, each pair of diagnostic probes is designed such that the probes hybridize immediately adjacent to each other on the same gDNA target in preparation for a subsequent ligation step. Each probe consisted of a target-specific sequence and either a common forward or reverse primer-specific sequence. One probe from each pair contained a "stuffer" sequence of 19–364 nucleotides (nt), which was unique and thus diagnostic for the particular probe pair. The LDLR MLPA kit contained 31 sets of probes, 16 of which were LDLR specific and the others were control standards. Reactions were carried out in 200 µl tubes in a model 9700 thermocycler (PE-Applied Biosystems). One hundred to 300 ng of gDNA from each subject was diluted in 5 µl of 1x Tris-EDTA and denatured at 98°C for 5 min. MLPA buffer and probe mix (1.5 µl of each) were then added to allow probes to anneal to target gDNA by heating at 95°C for 1 min and incubating at 60°C for 16 h. Annealed probes were ligated at 54°C for 15 min followed by inactivation at 98°C for 5 min. Ten microliters of ligation reaction was removed for multiplex amplification using a pair of common primers, of which one was labeled with the fluorescent dye FAM. Taq polymerase was added to the reaction at 60°C, followed by 33 cycles of 95°C for 30 s, 60°C for 30 s, 72°C for 1 min, and a final extension step of 72°C for 20 min. Two microliters of reaction solution was used for fragment analysis on the 3730 capillary sequencer, with LIZ-500 size standards (PE-Applied Biosystems). The procedure was performed according to the manufacturer's instructions (www.mrc-holland.com).
MLPA fragment analyses
Data analysis was performed using GeneMapper version 3.5 (PE-Applied Biosystems). Electropherograms of fragments from MLPA analysis of LDLR from normal subjects show a profile composed of 31 peaks (range, 130–382 nt). Exons 10 and 13 are not interrogated with the current version of the probe set. The multiplex contained five internal DNA quantity (DQ) control fragments. Four of these, of 64, 70, 76, and 82 nt, were ligation independent and were included to demonstrate that sufficient template DNA was present for the entire multiplex amplification reaction. The fifth DQ fragment, of 94 nt, indicated successful ligation by producing a peak of comparable size to that of the other chromosome-specific probes in the multiplex. The relative areas under the curve (AUCs) for peaks in each sample were determined. The relative peak AUC for each probe was calculated using four adjacent peaks, two on either side, as internal controls. The fraction of each peak was then divided by the median peak fraction of the corresponding fragment from five normal control samples. In 18 normal individuals, these calculations gave values close to 1.0, which corresponded to the normalized mean peak area and standard deviations for an individual with two copies of the target sequence. Copy number results >1.33 or <0.75 were flagged. Calculations were performed on samples processed within an assay run. Two-sample t-tests were used for all statistical comparisons against the profiles of normal samples, with nominal P < 0.05.
DNA sequence confirmation of deletion break point
As proof of the principle that a reduced MLPA peak in this study sample corresponded to a deleted region of gDNA, we identified five patients with a similar MLPA pattern, each showing 50% loss of height of exon 1. All of these patients had French surnames, leading us to consider that this particular MLPA pattern might have resulted from the >15 kb deletion (
>15 kb) at the 5' end of the LDLR. This deletion is the most common cause of FH in French Canadians (1, 13). We used the reported method (13) to sequence the deletion break points in these five patients. Briefly, primer pairs 23F (5'-CAC ATG CTC GAG TGT AAG AAC-3') and 17R (5'-GGT TGG CGT TTT TCA TAT TAG ACC-3') were used to amplify a 435 bp fragment spanning the deletion break point. When present, this fragment was amplified and sequenced. Primers 20F (5'-AGA CCC AGA AAG AAA TGA TTC G-3') and 20R (5'-GAA GTG TCC TTG TCT TCT CAT CG-3') produced a 236 bp fragment, which indicated the presence of the wild-type allele (13), confirming heterozygosity.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Mutations in FH patients found by EBESA
Twenty-one FH patients (30% of the sample) were heterozygous for a total of 18 different LDLR missense mutations, namely A29S, C42R, E80K, E119K, C163Y, D203N, E207K, G314S, Q363P, A370T (three patients), R395W, V408M, T413M, Y421C*, L458P, L561P (two patients), E760D, and N804K. Twelve FH patients (17.1% of the sample) were heterozygous for a total of seven different RNA splicing mutations, namely IVS2 donor +2T>G*; IVS3 donor +1G>A (two patients), IVS3 acceptor –2A>C, IVS7 acceptor –8T>C (three patients), IVS8 acceptor –10G>A, IVS12 donor +15C>A*, and IVS14 donor +5G>A (three patients). Two FH patients (3% of the sample) were each heterozygous for an in-frame deletion, namely Asp36 and Gly198*. Nine FH patients (12.9% of the sample) were heterozygous for a total of seven different nonsense mutations, namely W23X, fs R103P-X108, C152X, fs G197V-X243, fs F381-X418, fs K497L-X528*, and C660X (three patients). Among the mutations listed above are five newly discovered LDLR mutations, which are indicated by asterisks. All LDLR mutations were absent from the genomes of 100 healthy control subjects. Five FH patients (7.1% of the sample) had the APOB R3500Q mutation. Twenty-one FH patients (30% of the sample) had no LDLR mutation by EBESA and did not have the APOB R3500Q mutation.
MLPA analysis
The 21 FH patients who had neither a mutation in LDLR by EBESA nor the APOB R3500Q mutation were studied using LDLR MLPA. In addition, LDLR MLPA was performed on gDNA from 23 healthy normal subjects. LDLR deletion mutations were apparent on qualitative inspection of MLPA peak profiles. This is illustrated in Figs. 1, 2
, in which comparison of peak heights of a probe located within LDLR and neighboring control peaks indicated a gene dosage mutation. Normalized peak areas for exons were determined using the normal samples as described in Methods. Standard deviations of <7% of the normalized mean were recorded for all samples using the means of the relative copy numbers of deletion/duplication region probes. We set threshold values for the identification of abnormal results at 0.75 and 1.33 (corresponding to >4 standard deviations from the mean) (Table 1). Among the 21 previously mutation-free FH patients, there were seven abnormal MLPA patterns seen in 12 patients. Four involved reduced peaks for one exon only: exon 1 (five patients), exon 2, exon 3, and exon 6. Three involved reduced peaks for consecutive exons: exon 1–6 (two patients), exon 2–6, and exon 4–14. Control MLPA profiles of three FH patients who had missense mutations in LDLR were all normal (data not shown).
|
|
|
|
exon 1 each had French surnames. This led us to consider that this abnormal MLPA pattern might have been the result of the >15 kb deletion at the 5' end of the LDLR that is the most prevalent cause of FH in French-Canadian patients (13). We used the reported method to detect >15 kb in these five patients (13). We first found that each had the diagnostic 435 bp band that could only be amplified if the deletion was present (13), that each was heterozygous for this mutant fragment, and that the sequence of this fragment corresponded to >15 kb (13) in all five subjects (Fig. 3)
.
Biochemical attributes according to molecular diagnosis
We stratified patients into the following molecular subgroups: 1) heterozygotes for missense mutations; 2) heterozygotes for splicing mutations; 3) heterozygotes for either nonsense mutations or in-frame deletions; 4) abnormal MLPA pattern; 5) heterozygotes for APOB R3500Q; and 6) no LDLR mutation and normal APOB receptor binding domain sequence (Table 1). The mean LDL cholesterol among the small group of 12 FH patients with abnormal LDLR MLPA patterns was not different from that of the other 58 patients (7.98 ± 2.36 mmol/l versus 6.98 ± 1.47 mmol/l; P = 0.19 by Student's t-test using unequal variances).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
>15 kb, at the 5' end of LDLR. Although LDLR missense mutations are most common in FH, MLPA abnormalities, mainly deletions, also appear to be prevalent. These results indicate that MLPA might be a promising method to detect additional gDNA abnormalities of the LDLR gene. Technology is an important determinant of the mutation type detected. For instance, in the pre-PCR era, the most prevalent LDLR mutation type in FH was large-scale gDNA rearrangement, such as large insertions and deletions (14). One reason for this was the relative enrichment within LDLR of sequences that were predisposed to imperfect recombination events or exon shuffling (14). But another reason was the fact that most LDLR mutations in the late 1980s were detected by Southern blotting. Southern blots were ideal for detecting large-scale gDNA rearrangements involving >200 nt of target gDNA, which was the resolution limit for DNA fragments. However, detection of small mutations was beyond the Southern blot's resolving capacity, unless a mutation altered a recognition site for a restriction endonuclease. Because Southern blotting was optimal for detecting large-scale LDLR rearrangements, by the early 1990s these represented a large proportion of FH mutations (1, 14).
PCR made it easier to quickly detect small mutations from gDNA using relatively few resources and with less expense, effort, and radiation requirements than Southern blotting. Thus, most of the 1,000 LDLR mutations in FH that are in current databases are of the single-nucleotide variety discovered by exon-by-exon PCR-based screening methods such as EBESA (http://www.ucl.ac.uk/fh/). However, PCR-based methods of interrogating gDNA are practically limited to regions of 200–700 bp. A large deletion affecting one LDLR allele may never be detected by PCR, because there is no target sequence upon which amplification primers can anneal. Only the normal allele is amplified, creating effective hemizygosity for that region of gDNA. Analysis of the products of such a reaction in the search for a DNA change, whether indirectly (15) or by EBESA, will yield a false-negative result.
MLPA optimally detects alterations over a relatively large gDNA region but uses no radioactivity and is PCR based. As a semiquantitative method, internal standards and controls are required for each analysis, but this need is met by the high throughput afforded by rapid, automated fragment detection. Our cost for LDLR MLPA analysis per patient sample, including reagents, duplicate analyses, controls, and labor, was $120 US, which seems cost-effective considering the increment in molecular diagnosis. To date there is no optimal procedure for data normalization and for detection of abnormal data points for MLPA (16). For future clinical applications, performance of LDLR MLPA will need to be assessed by blinded analysis of FH samples that have well-defined gDNA rearrangements. Various reference samples, including healthy controls and subjects with other mutation types, will need to be assessed, as will intraobserver and interobserver variability. It was promising that we sequenced the deletion break points at the 5' end of LDLR in five subjects whose MLPA analysis showed the exon 1 pattern.
MLPA reduced the percentage of FH patients who had no apparent abnormality in gDNA to 12.9% (9 of 70). What is the basis of the FH phenotype in the remaining subjects? Some of these patients might have had other LDLR mutation types, such as mutations deep within introns or beyond the region of the promoter and 3'-untranslated regions that could be accessed by our EBESA and MLPA primers. Alternatively, some patients may have had another genetic cause for their FH phenotype, as exemplified by the five patients with APOB R3500Q. We used EBESA to examine PCSK9 (MIM 607786) and ARH (MIM 605747) in the nine FH patients with no LDLR mutation and normal APOB exon 26 sequence and found no mutations in these genes (data not shown). Mutations in PCSK9 and ARH might still be found in some of these FH patients through MLPA. Alternatively, causative mutations in some of these FH patients may be found in other genes.
In summary, we report abnormal MLPA patterns in a substantial proportion of FH patients who had neither a LDLR mutation using EBESA nor the APOB R3500Q mutation. MLPA increased the LDLR mutation detection rate in FH subjects with no APOB R3500Q mutation from 67.6% to 86.2% (from 44 of 65 to 56 of 65). If this increment in mutation detection can be replicated, MLPA might become a useful adjunct to existing screening methods for the molecular diagnosis of FH. Before MLPA can be applied clinically, it requires formal evaluation of such performance criteria as sensitivity, specificity, and reproducibility. However, the clinical importance of a proper FH diagnosis (2) supports the need to develop sensitive molecular diagnostic algorithms to optimally diagnose FH. The data reported here also suggest the possibility of additional molecular heterogeneity of FH.
|
ACKNOWLEDGMENTS |
|---|
Manuscript received October 21, 2004
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  Z. Awan, K. Alrasadi, G.A. Francis, R.A. Hegele, R. McPherson, J. Frohlich, D. Valenti, B. de Varennes, M. Marcil, C. Gagne, et al. Vascular Calcifications in Homozygote Familial Hypercholesterolemia Arterioscler. Thromb. Vasc. Biol., April 1, 2008; 28(4): 777 - 785. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. L. Pollex and R. A. Hegele Copy Number Variation in the Human Genome and Its Implications for Cardiovascular Disease Circulation, June 19, 2007; 115(24): 3130 - 3138. [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Hegele Copy-number variations add a new layer of complexity in the human genome Can. Med. Assoc. J., February 13, 2007; 176(4): 441 - 442. [Full Text] [PDF]
|
|
|
|
 |
|
 O. W. Souverein, J. C. Defesche, A. H. Zwinderman, J. J.P. Kastelein, and M. W.T. Tanck Influence of LDL-receptor mutation type on age at first cardiovascular event in patients with familial hypercholesterolaemia Eur. Heart J., February 1, 2007; 28(3): 299 - 304. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. S. Sullivan, S. J. Bowne, C. R. Seaman, S. H. Blanton, R. A. Lewis, J. R. Heckenlively, D. G. Birch, D. Hughbanks-Wheaton, and S. P. Daiger Genomic Rearrangements of the PRPF31 Gene Account for 2.5% of Autosomal Dominant Retinitis Pigmentosa. Invest. Ophthalmol. Vis. Sci., October 1, 2006; 47(10): 4579 - 4588. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Journal of Biological Chemistry |
| Molecular and Cellular Proteomics | ASBMB Today |
