Download article in press - MRC

ATH-9692;
ARTICLE IN PRESS
No. of Pages 10
Atherosclerosis xxx (2006) xxx–xxx
Genetic defects causing familial hypercholesterolaemia: Identification of
deletions and duplications in the LDL-receptor gene and summary
of all mutations found in patients attending the
Hammersmith Hospital Lipid Clinic
Isabella Tosi a,1 , Paola Toledo-Leiva a,1 , Clare Neuwirth a,b ,
Rossi P. Naoumova a,b , Anne K. Soutar a,∗
b
a MRC Clinical Sciences Centre, Imperial College London, United Kindom
Lipid Clinic, Hammersmith Hospital, Du Cane Road, London W12 0NN, United Kingdom
Received 3 July 2006; received in revised form 2 October 2006; accepted 6 October 2006
Abstract
Familial hypercholesterolaemia (FH) results from defective catabolism of low density lipoproteins (LDL), leading to premature atherosclerosis and early coronary heart disease. It is commonly caused by mutations in LDLR, encoding the LDL receptor that mediates hepatic uptake
of LDL, or in APOB, encoding its major ligand. More rarely, dominant mutations in PCSK9 or recessive mutations in LDLRAP1 (ARH)
cause FH, gene defects that also affect the LDL-receptor pathway. We have used multiplex ligation-dependent probe amplification (MLPA)
to identify deletions and rearrangements in LDLR, some not detectable by Southern blotting, thus completing our screening for mutations
causing FH in a group of FH patients referred to a Lipid Clinic in London. To summarise, mutations in LDLR were found in 153 unrelated
heterozygous FH patients and 24 homozygotes/compound heterozygotes, and in over 200 relatives of 80 index patients. LDLR mutations
included 85 different point mutations (7 not previously described) and 13 different large rearrangements. The APOB R3500Q mutation was
present in 14 heterozygous patients and a mutation in PCSK9 in another 4; LDLRAP1 mutations were found in 4 “homozygous” FH patients.
Our data confirm that DNA-based diagnosis provides information that is important for management of FH in a considerable number of families.
© 2006 Elsevier Ireland Ltd. All rights reserved.
Keywords: Plasma cholesterol; MLPA; LDLRAP1; PCSK9; APOB; Heterozygous FH; Homozygous FH
1. Introduction
Familial hypercholesterolaemia (FH) is an autosomal
dominant disorder caused by defective clearance of low
density lipoproteins (LDL) from the circulation, leading to
premature atherosclerosis and a marked increased risk of
coronary heart disease. FH is usually caused by mutations
in the gene for the LDL receptor (LDLR); a mutation in the
gene for the ligand for the LDL receptor, apolipoprotein B100
(ApoB100), results in the same phenotype, but this disorder
∗
1
Corresponding author. Tel.: +44 20 8383 2324; fax: +44 20 8383 2028.
E-mail address: [email protected] (A.K. Soutar).
These authors contributed equally to this work.
is often referred to as familial defective ApoB100 (FDB)
[1].
During various studies in which we have investigated the
genetic defect underlying a firm clinical diagnosis of definite heterozygous FH in patients in the UK, we have failed to
detect a mutation in the genes for the LDL receptor (LDLR) or
apolipoprotein B (APOB) in up to 15% of any group. At first,
this was ascribed to lack of sensitivity in the analysis methods used, but the advent of high quality, automated nucleotide
sequencing of PCR-amplified DNA has not greatly reduced
the number of patients with an unidentified mutation, suggesting that some defects in the LDLR remain undetected or
that other genes may be involved. The identification of dominant mutations in PCSK9 as a cause of heterozygous FH
0021-9150/$ – see front matter © 2006 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.atherosclerosis.2006.10.003
Please cite this article in press as: Tosi I et al., Genetic defects causing familial hypercholesterolaemia: Identification of deletions and
duplications in the LDL-receptor gene and summary of all mutations found in patients attending the Hammersmith Hospital Lipid Clinic,
Atherosclerosis (2006), doi:10.1016/j.atherosclerosis.2006.10.003
ATH-9692;
No. of Pages 10
2
ARTICLE IN PRESS
I. Tosi et al. / Atherosclerosis xxx (2006) xxx–xxx
have revealed that defects in other genes can result in this
phenotype, but these remain a rare cause of FH [2].
In order to define a group of patients that warrant further investigation of the underlying cause of their inherited
hypercholesterolaemia, it was necessary to exclude those
with a large deletion or duplication in the LDLR. Although
Southern blotting has been used successfully for this purpose by ourselves and others [3–7] it is time-consuming,
technically demanding, and requires large amounts of high
quality DNA. Most importantly, it lacks the sensitivity to
detect all rearrangements, for example deletions encompassing most of the coding region of the gene so that a cDNA
probe cannot readily detect abnormal restriction enzyme
fragments.
Recently, a novel method for the detection of gene deletions and duplications has been devised: multiplex ligationdependent probe amplification (MLPA) [8]. The method
depends on the hybridisation of two short specific oligonucleotide probes to abutting regions of an exon in genomic
DNA from a patient. If the exon is intact, the two oligonucleotides can be ligated and the resultant fragment amplified by PCR with fluorescently labelled primers for which
sequences are incorporated at the end of each probe. Pairs
of probes for many exons of one or more genes, all containing this same pair of universal primer sequences, can be
hybridised, ligated and amplified in single multiplex reactions. The size of each amplified product is determined by
the inclusion of a “filler” sequence of different length in each
probe and thus the PCR products can readily be analysed on
a DNA sequencer. Comparison of the relative amount of each
product for the gene in question with that from control genes
on different chromosomes allows the identification of exons
that are absent (homozygous deletion) or present at half copy
number (heterozygous deletion) or in multiple copy number
(heterozygous duplication) in a patient. Here we describe its
use to detect 12 different gene rearrangements in the LDLR,
several of which would not be detected readily by Southern
blotting. The majority of these have been confirmed by PCRamplification across the deletion/duplication joint in DNA or
mRNA and/or by their presence in affected relatives.
characterised by extreme hypercholesterolaemia with serum
cholesterol between 14 and 30 mmol/l, onset of cutaneous
planar or tuberous xanthomas in early childhood plus tendon xanthomas and corneal arcus. The patients were either
referred to the Lipid Clinic or their samples were sent to us for
molecular diagnosis. All patients had given informed consent
to DNA-based diagnosis of their disorder.
2.2. Identification of point mutations
Methods used to identify point mutations and minor insertion deletions in LDLR, APOB and PCSK9 by sequencing of
amplified fragments of genomic DNA or mRNA from immortalized lymphocytes have been described in detail elsewhere
[10–12].
2.3. MLPA
DNA samples from patients were analysed by MLPA
according to the supplier’s instructions [13] and compared
with at least five control samples analysed at the same
time. The PCR products were fractionated on an ABI DNA
sequencer and the data analysed with GenotyperTM software.
Only electropherograms that passed quality control were
analysed, i.e. the peak heights of non-ligated probes were
negligible compared with ligated probe fragments. Peaks on
each electropherogram were normalised by expressing their
height as a fraction of the total height of all control peaks.
Relative amounts of each LDLR exon in each patient sample
were determined by dividing the normalised value for each
peak by the average normalised value for that amplicon from
at least five controls (variance <10%). Exons present at less
than 0.75-fold the control value were deemed to have half the
normal copy number, and those at more than 1.3-fold to be
duplicated.
Amplification of genomic DNA to confirm the presence
of deletions was carried as described previously [10]; details
of the primers and the conditions used are shown in Table 1.
3. Results
2. Methods
2.1. Patients
The heterozygous FH patients whose DNA was analysed
in this study had all attended the Hammersmith Hospital
Lipid Clinic during the last 10–15 years. The diagnosis of
FH was based on clinical criteria established by the Simon
Broome Study Group [9], i.e. total plasma cholesterol concentration greater than 7.5 mmol/l in the proband, together
with either tendon xanthoma in the proband or in a first degree
relative (definite FH), or the presence of premature coronary heart disease or hypercholesterolaemia in a first degree
relative (possible FH). The homozygous FH patients were
Table 2 shows the different mutations that have been identified in 145 apparently unrelated heterozygous FH probands
attending the Hammersmith Hospital Lipid Clinic over the
last 10–15 years, while Table 3 summarises the genetic
defects found in 28 patients with a clinical diagnosis of suspected homozygous FH who were either referred to the clinic
or whose samples were sent to us for genetic characterisation of the underlying defect ([6,10,14–22] and unpublished
data). Several additional mutations have also been identified
by us in a group of heterozygous FH patients from Cardiff
[10], and in both homozygous and heterozygous patients of
Chinese origin [23,24].
Patients in whom no defect could be identified by sequencing were analysed for major deletions and insertions in
Please cite this article in press as: Tosi I et al., Genetic defects causing familial hypercholesterolaemia: Identification of deletions and
duplications in the LDL-receptor gene and summary of all mutations found in patients attending the Hammersmith Hospital Lipid Clinic,
Atherosclerosis (2006), doi:10.1016/j.atherosclerosis.2006.10.003
ATH-9692;
No. of Pages 10
ARTICLE IN PRESS
I. Tosi et al. / Atherosclerosis xxx (2006) xxx–xxx
3
Table 1
PCR primers and conditions for confirmation of rearrangements in LDLR
LDLR mutation
Primers: name
Sequence
Conditions for mutant allele
Del e5
F: AKS54; R: 6R
5 -CCCCAGCTGTGGGCCTGCGACA-3 ;
Del e13–14
F: AKS230; R: AKS233
Del e16–17
F: 15F; R: 18R
Dup e9–14
F: AKS232; R: AKS227
Dup e11–12
F: 12F; R: 11R
94 ◦ C × 2 ; 94 ◦ C × 30 , 62 ◦ C × 45 ,
68 ◦ C × 1 /30 cycles; 68 ◦ C × 8
94 ◦ C × 2 ; 94 ◦ C × 1 , 65 ◦ C × 3 ,
68 ◦ C × 1 /35 cycles; 68 ◦ C × 8
94 ◦ C × 2 ; 94 ◦ C × 30 , 59 ◦ C × 45 ,
68 ◦ C × 4 /35 cycles; 68 ◦ C × 8
94 ◦ C × 2 ; 94 ◦ C × 30 , 61 ◦ C × 45 ,
68 ◦ C × 1 /30 cycles; 68 ◦ C × 8
94 ◦ C × 2 ; 94 ◦ C × 30 , 61 ◦ C × 45 ,
68 ◦ C × 1 /30 cycles; 68 ◦ C × 8
5 -GCAGAGTGGAGTTCCCAAAACC-3
5 -CCGCCTCTACTGGGTTGACTCCAAACTTCAC-3 ;
5 -GCTGACCTTTAGCCTGACGGTGGATG-3
5 -CCAAGGTCATTTGAGACTTTCGTCA-3 ;
5 -TGGTGCCATCTGCTGTTGTGT-3
5 -AGAGGACCACCCTGAGCAATGGCGG-3 ;
5 -GCGACCACGTTCCTCAGGTTGGGGATGAGG-3
5 -GGTGCTTTCTGCTAGGTCC-3 ;
5 -AGCAGCTTGGGCTTGTCCCAGA-3
Del, deletion; Dup, duplication; e, exon; F, forward primer; R, reverse primer; bp, base pair.
LDLR by MLPA. DNA samples from 50 unrelated patients
with a diagnosis of definite (N = 32) or possible (N = 18)
heterozygous FH were analysed; of these, 14 were found
to have a deletion or duplication of one or more exons
in the LDLR. These data are summarised in Fig. 1 and
Table 4, and points of interest described in more detail
below. Panel A in Fig. 1 (Fig. 1A) shows the results
from four patients in whom no major rearrangement was
found.
3.1. Deletions of single exons
DNA from two unrelated patients was found to have
reduced copy number of exon 8 (Fig. 1B, samples 629
and 663); however, amplification and sequencing of exon
8 from genomic DNA revealed that both these individuals
were heterozygous for a 4 bp duplication in this exon (bases
GTGG1118–1121 , where base 1 is A of the ATG initiator
codon) that encompassed the region of the exon where the
Fig. 1. Results of MLPA analysis of the LDL-receptor gene (LDLR) in genomic DNA from FH patients. Peak heights for each PCR product were normalised
relative to the control probes within each sample and then expressed relative to the mean peak height of each in at least five control DNA samples run at the
same time. The x-axis represents the size of the PCR product from which the identity of the product is determined. (A) Representative MLPA patterns from FH
patients in whom no rearrangements were detected. (B) Patients with small deletions of one or two exons at the 5 end of LDLR; ** indicates the peak obtained
with the probe to SMARCA1, the control gene adjacent to the 5 end of the LDL receptor gene (see Fig. 3); ‡ indicates a control gene that is present at half
copy number in two subjects. (C) patients with duplications in LDLR. (D) Patients with large deletions or deletions at the 3 end of LDLR; * indicates the peak
obtained with the probe to ANKRD25, the control gene in the region flanking the 3 end of LDLR.
Please cite this article in press as: Tosi I et al., Genetic defects causing familial hypercholesterolaemia: Identification of deletions and
duplications in the LDL-receptor gene and summary of all mutations found in patients attending the Hammersmith Hospital Lipid Clinic,
Atherosclerosis (2006), doi:10.1016/j.atherosclerosis.2006.10.003
ATH-9692;
ARTICLE IN PRESS
No. of Pages 10
4
I. Tosi et al. / Atherosclerosis xxx (2006) xxx–xxx
Table 2
LDLR mutations identified in 153 unrelated heterozygous FH patients attending Hammersmith Hospital Lipid Clinic
LDLR mutation
Exon
No. of index patients
M-21L (nt A1 → T)
Q12X (nt C97 → T)
C25G (nt T136 → G)a
C47X (nt C204 → A)a
del nt GT196–197
C54W (nt T225 → G)a
W66G (nt T260 → G)
W66X (nt G261 → A)
C68Y (nt G266 → A)
D69G (nt A269 → C)
E80K (nt G301 → A)
del nt G303
nt C313+1 G → A
nt C313+5 G → A
C88Y (nt G326 → A
E119D (nt G420 → C)
C146X (C501 → A)
Y167X (nt C564 → G)
del G197 (nt GGT652−4 )
D200N (nt G661 → A)
D200G (nt A662 → G)
dup 21 nt663–683
D206E (nt C681 → G)
del nt AC680−1
E207Q (nt G682 → C)
E207X (nt G682 → T)
C210X (nt C693 → A)
ins C after nt C704
C227Y (nt G743 → A)
S265R (nt C858 → A)
D280A (nt A902 → C)
D283N (nt G910 → A)
D283E (nt C912 → G)
D286H (nt G919 → C)
C292Y (nt G938 → A)
C292X (nt C939 → A)
R329X (nt C1048 → T)
R329P (nt C1048 → G)
del nt C1008−31
ins GTGG at nt T1112
C356Y (nt G1130 → A)
C358R (ntT1135 → C)
Q363X/D365E (nt
C1150 → T; C1158 → G)
C371X (nt C1176 → A)
nt C1216 A (splice)
R385Q (nt G1217 → A)
E387K (nt G1222 → A)
Y398X (nt C1257 → G)
V408M (nt G1285 → A)
del nt G1358+55 , ins CGGCT
Q434X (nt C1363 → T)a
L458P (nt T1436 → C)
D461H (nt G1444 → C)
D461N (nt G1444 → A)
del CT at (nt C1477−8 )
D471N (nt G1474 → A)
V502M (nt G1567 → A)
A519T (nt G1618 → A)
Q540X (nt C1681 → T)
G544A (nt G1694 → C)
L578S (nt T1786 → C)
ex 1
ex 2
ex 2
ex 2
ex 3
ex 3
ex 3
ex 3
ex 3
ex 3
ex 3
ex 3
intron 3
intron 3
ex 4
ex 4
ex 4
ex 4
ex 4
ex 4
ex 4
ex 4
ex 4
ex 4
ex 4
ex 4
ex 4
ex 5
ex 5
ex 6
ex 6
ex 6
ex 6
ex 6
ex 6
ex 6
ex 7
ex 7
ex 7
ex 8
ex 8
ex 8
ex 8
1
1
1
1
1
1
5
1
4
3
4
1
5
1
1
1
1
1
8
2
3
1
2
6
1
6
1
1
1
1
1
1
1
1
2
2
4
1
1
1
1
1
1
ex 8
ex 9
ex 9
ex 9
ex 9
ex 9
intron 9
ex 10
ex 10
ex 10
ex 10
ex 10
ex 10
ex 10
ex 11
ex 11
ex 11
ex 12
1
1
1
2
1
2
3
1
2
3
3
1
1
1
1
1
1
1
Table 2 (Continued )
LDLR mutation
Exon
No. of index patients
P587R (nt C1828 →
nt G1845+11 C → G)
F598L (nt T1855 → C)
W599R (nt T1858 → C)
A606D (nt C1899 → A)
R612C (nt C1897 → T)
P628L (nt C1946 → T)
del G1987
C660Y (nt C2042 → A)
C656R (nt T2029 → C)
P664L (nt C2054 → T)
ins C at nt2061
A676P (nt G2089 → C)a
R723Q (nt2231 → A)
ex 12
intron 12
ex 13
ex 13
ex 13
ex 13
ex 13
ex 13
ex 14
ex 14
ex 14
ex 14
ex 14
ex 15
1
1
1
1
1
3
1
1
1
1
8
1
1
1
Deletion
Deletion
Deletion
Deletiona
Deletiona
Deletion
Deletion
Deletion
Deletion
Deletion
duplication
duplication
Deletion
ex 1
ex 2–5
ex 2–6
ex 2–18
ex 1–18
ex 5
ex7
ex 7–8
ex 7–14
ex 13–14
ex 9–14
ex 11–12
ex 16–17
1
1
1
1
1
5
1
1
1
2
1
1
1
a Mutations not listed on LDLR mutation databases (http://www.ucl.
ac.uk/fh/ and http://www.umd.necker.fr).
probes annealed. Careful inspection of the MLPA results
shows that the apparent copy number of exon 8 is more
than half, perhaps suggesting that some amplification of the
mutant allele occurred.
DNA from two other patients was found to have half
the normal copy number of exon 5 (Fig. 1B, samples 567
and 617); in both cases, deletion of exon 5 was confirmed
by PCR amplification of genomic DNA across the deletion
joint in the probands and in affected relatives. As shown
in Fig. 2A for proband 567 and his hypercholesterolaemic
daughter, amplification of genomic DNA with a forward
primer located in exon 4 and a reverse primer in exon 6 produced the expected 2.2 kb normal product with DNA from the
proband, his daughter and a control subject, together with an
additional product of approximately 1.3 kb in the proband and
his daughter. This mutation has been observed previously in
the UK [6].
DNA from one patient (547 in Fig. 1B) had half the normal copy number of exon 1; the same deletion was found in
DNA from the patient’s affected father by MLPA (data not
shown). Sequencing of the promoter and exon 1 of the LDLR
in amplified genomic DNA from the patient did not reveal
any variation that could explain the apparent deletion of this
exon, but the deletion joint could not be amplified with any of
several primer sets tested. This is probably because intron 1 is
very large and the deletion at the 5 end could have extended
27 kb upstream as far as the adjacent gene, SMARCA4 (Fig. 3)
which was intact in the patient’s DNA (probe ** in Fig. 1B).
Please cite this article in press as: Tosi I et al., Genetic defects causing familial hypercholesterolaemia: Identification of deletions and
duplications in the LDL-receptor gene and summary of all mutations found in patients attending the Hammersmith Hospital Lipid Clinic,
Atherosclerosis (2006), doi:10.1016/j.atherosclerosis.2006.10.003
ATH-9692;
ARTICLE IN PRESS
No. of Pages 10
I. Tosi et al. / Atherosclerosis xxx (2006) xxx–xxx
5
Table 3
LDLR mutations identified in 28 “homozygous” FH patients referred to Hammersmith Lipid Clinic or to Lipoprotein Group for genetic characterisation
Gene
Genotype
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
LDLR
AB
AB
AB
AB
AB
AB
AB
AB
ABb
AB
AB
AB
AB
AB
AAc
AA
AAc
AA
AA
AA
AAc
AAc
AAc
AAc
LDLRAP1f
LDLRAP1
LDLRAP1
LDLRAP1
AB
AAc
AAc
AA
a
b
c
d
e
f
Mutation(s)
Allele A
Allele B
L578S
E80K
C227Y
P664L
D69G
D461N
Del ex 13–15a
C176Ra (nt T588 → C)
P664Lb
C68Y
F220Sa (nt T722 → C)
D280A
E80K
P664L
P664L
R385Pa (nt G1217 → C)d
E387K
C292X
C292X
Q540X
W-18Xa (nt G11 → A)
C281Wa (nt C906 → G)
G528Da (nt G1646 → A)
D112Na (nt G397 → A)d
del ex 2–6
del exon 1–6a
R329P
Unknown
D283E
dup nt 663–683
R612C
P664L
P664Lb
del ex 16–17
E207X
S265R
del nt A2292 a
del G736 a
–
–
–
–
–
–
–
–
–
–
UK
UK
UK
UK
UK
UK
UK
UK
UK
UK
UK
UK/Greek (B)
UK/Irish (B)
Asian Indian
Asian Indian
Asian Indian
Asian Indian
Greek Cypriote
Greek Cypriote
AfroCaribbean
Colombian
Iraqi
Kosovan
Omani
English
Turkish (Lebanese)
Pakistani
Sardinian
Mutation not found in heterozygous FH patients.
Same mutation on different haplotype inherited from unrelated parents [21].
Parents known to be consanguineous.
Mutations not described previously (see Table 2).
Patients not known to be related, but common Greek mutation [25].
Formerly known as ARH, for autosomal recessive hypercholesterolaemia [26].
3.2. Deletions of two or more exons
As shown in Fig. 1B, DNA from patient 614 was found to
half the normal copy number of exons 7 and 8 of the LDLR,
Table 4
Deletions and rearrangements in LDLR detected by MLPA
Patient
DNA no.
Major
rearrangement
Confirmation
Affected relative
PCR
567
659
617
549
580
564
565
629
547
614
560
397
556
Del ex 5
Del ex 1–18
Del ex 5
Dup ex 11–12
Del ex 13–14
Del ex 2–18
Del 2–6
Del ex 8 (4 bp del)
Del ex 1
Del ex 7–8
Dup ex 9–14
Del ex 2–5
Del ex 16–17
Y
Na
Y
Y
N
Y
Y
Y
Y
Y/N
Y
Y
Y
PCR
–
PCR
PCR
PCR
–
–
DNA sequence
–
PCR
PCR
mRNA sequence
PCR
a
Ethnic origin
Observed in two independent samples of genomic DNA.
and apparently also half the normal copy number of a control
gene locus on chromosome 10p14 (probe ‡) which encodes a
sequence 11 kb from the 10p telomere encoding the hypothetical gene LOC254312 [13]. Individual 614 was the sibling
of the proband in our study, but the data revealed that the
proband did not carry the deletion of exons 7 and 8, but had
inherited the variant allele of chromosome 10 (Fig. 1B). No
other DNA samples tested, including controls, were found to
carry this variant, and we conclude that it occurs with an allele
frequency of less than ∼0.005. Since this locus encodes a
hypothetical gene for which there is no evidence that it might
be associated with lipid metabolism, this gene variant was
not investigated further.
DNA from one patient (659 in Fig. 1D) was found to have
half the normal copy number of exons 1–18, and since the
entire coding region of LDLR was apparently deleted from
one allele, this deletion would not be detected by Southern
blotting with cDNA probes. No affected relatives were available from this subject, but the deletion was observed in two
separate DNA samples from the same individual. The full
extent of this deletion could not be determined by PCR, but
Please cite this article in press as: Tosi I et al., Genetic defects causing familial hypercholesterolaemia: Identification of deletions and
duplications in the LDL-receptor gene and summary of all mutations found in patients attending the Hammersmith Hospital Lipid Clinic,
Atherosclerosis (2006), doi:10.1016/j.atherosclerosis.2006.10.003
ATH-9692;
6
No. of Pages 10
ARTICLE IN PRESS
I. Tosi et al. / Atherosclerosis xxx (2006) xxx–xxx
Fig. 2. Confirmation of deletions and duplications by PCR. (A) Genomic DNA from patient 567 (Pr), his daughter (Da) and a control subject (Co) was amplified
with the indicated primers and the products analysed on a 1% agarose gel stained with ethidium bromide; the diagram above shows part of LDLR with the
deleted exons shaded. (B) mRNA from immortalised B-lymphocytes from patient 397 (Pr) and a control subject (Co) was amplified by RT-PCR and sequenced
with a primer from nt 1137–1087 in the cDNA, where 1 is the A of the ATG initiator codon; the sequence shown is the non-coding strand. (C) Genomic DNA
from patient 580 (Pr) and a control subject (Co), was amplified and analysed as described in A above. (D) Amplified genomic DNA from patient 556 (Pr), his
sibling (Si), father (Fa) and mother (Mo). (E) Amplified genomic DNA from patient 560 (Pr), her twin (Tw), another sibling (Si, patient 710 in Fig. 1) and a
control subject (Co); Bl, water blank. (F) Amplified DNA from patient 549 (Pr) his uncle (Un, patient 713 in Fig. 1) and two control subjects (Co); Bl, water
blank.
both flanking control genes were present in apparently normal
copy number (probes SMARCA4** and ANKRD25*, 33 kb
downstream; see Fig. 3). DNA from another patient (564 in
Fig. 1D) had half the normal copy number of exons 2–18;
again the deletion joint could not be amplified but the downstream control gene (probe *) was present in normal copy
number. Two further affected members of the same large fam-
Fig. 3. Location of control genes adjacent to the LDL receptor. The
location of the control genes was determined from the ENSEMBL website: http://www.ensembl.org/Homo sapiens/contigview?c=19:11082496.5;
w=500000.
ily severely affected by FH, the proband’s niece and a sibling,
were found to have this same deletion by MLPA analysis.
Again, this deletion was not detectable by Southern blotting
with cDNA probes.
DNA from patient 565 had half the normal copy number of exons 2–6 (Fig. 1D); this mutation has been observed
previously in FH patients in the UK [20]. MLPA analysis
showed that the affected mother of this patient carried the
same deletion, but his son did not (data not shown).
DNA from patient 397 had half the copy number of exons
2–5 (Fig. 1D) and her affected daughter was found to have
the same deletion (data not shown). In an immortalised lymphocyte cell line from proband 397, the sequence of LDLR
mRNA amplified by RT-PCR confirmed the deletion of exons
2–5 (Fig. 2B).
DNA from one patient was found to have half the normal
copy number of exons 13 and 14, and deletion of these two
Please cite this article in press as: Tosi I et al., Genetic defects causing familial hypercholesterolaemia: Identification of deletions and
duplications in the LDL-receptor gene and summary of all mutations found in patients attending the Hammersmith Hospital Lipid Clinic,
Atherosclerosis (2006), doi:10.1016/j.atherosclerosis.2006.10.003
ATH-9692;
No. of Pages 10
ARTICLE IN PRESS
I. Tosi et al. / Atherosclerosis xxx (2006) xxx–xxx
7
Table 5
Genetic variation in PCSK9 in patient without known mutations
Genetic variation in PCSK9
Number of individuals
Location
Sequence variant
Hmz common
Htz/Hmz rare
Exon 1
Exon 1
Intron 4
Intron 4
Intron 5
Exon 9
Exon 9
15–16 ins L(+L) ins CTG at nt 287
R46L exon nt G381 → T
Exon 4 nt C + 4 → T
Exon 5 nt C − 7 → T
Exon 5 nt A + 3 → G
V460V nt G1624 → A
I474V nt A1664 → G
20
23
31
22
24
29
26
5/0
1/1
1/0
10/0
8/0
3/0
6/0
exons from genomic DNA was confirmed by PCR across the
deletion joint with a forward primer in exon 12 and a reverse
primer in exon 15. As shown in Fig. 2C, the normal allele
was not amplified (expected product of 6.3 kb) in DNA from
either the proband or unaffected controls, but a 2.7 kb product
was observed in DNA from the proband. Deletion of these
two exons has also been observed previously in an FH patient
in the UK population [27].
Finally, DNA from one patient was found to have half the
normal copy number of exons 16 and 17, a deletion that was
also confirmed by amplification across the deletion joint in
genomic DNA from the patient and his mother. As shown
in Fig. 2D, with PCR primers located in exons 15 (forward)
and 18 (reverse), the expected normal allele of 8.6 kb could
not be amplified with the conditions used, but a product of
approx. 4.8 kb was obtained with DNA from the proband and
his mother that was not observed with DNA from control
subjects.
3.3. Duplications
Two probands were found to have duplications of part of
the LDLR. In one patient (549 in Fig. 1C), more than two
copy numbers of exons 11 and 12 were present; this was also
observed in his maternal uncle (713 in Fig. 1C). Amplification of genomic DNA with a forward primer in exon 12 and
a reverse primer in exon 11 produced a product of approx.
2.0 kb with DNA from the proband and his uncle, but not
with DNA from an unaffected control subject (Fig. 2F). This
confirmed the presence of a tandem duplication of exons 11
and 12 in intron 12.
Another patient (560 in Fig. 1C) was found to have multiple copies of exons 9–14, and this was also observed in her
twin sister and a further sister (710 in Fig. 1C). Interestingly,
in all three cases and in several separate MLPA runs, the copy
number appeared to be greater than could be explained by a
single duplication of these exons. As can be seen in Fig. 1C,
all the duplicated peaks were present at approx. 3-fold the
normalised level, compared with the expected 1.5-fold that
is observed when there are three copy numbers (for example,
see the duplication of exons 11 and 12 described above; DNA
713 and 549 in Fig. 1C). The presence of at least one tandem
duplication of exons 9–14 was confirmed by PCR with a forward primer located in exon 14 and a reverse primer in exon
9; a product of approx. 2 kb was obtained with DNA from
the proband, her affected twin and a second affected sister
that was not obtained with DNA from a control individual
(Fig. 2E). This confirmed the presence of a duplication of
exons 9–14 in intron 14, but could not distinguish between
one or more duplications of this gene segment.
3.4. Screening for mutations in APOB and PCSK9
Genomic DNA from 32/35 patients with no detectable
mutation was screened for possible rare mutations in APOB
and PCSK9; DNA was no longer available from the other
three. For APOB, codons 3431–3584 (in exon 26) and
4310–4396 (in exon 29) were amplified and sequenced, but
the only sequence variant found was the common polymorphism in codon N4311 [28]; 7 (22%) patients were heterozygous and 4 (12%) were homozygous for the less frequent G
allele, while the remainder were homozygous for the more
frequent A allele. For PCSK9, the coding exons and flanking
sequences were sequenced; only known polymorphisms that
have been detected in normolipaemic individuals were found
[29], as shown in Table 5.
4. Discussion
Out of 206 unrelated heterozygous FH patients who have
now been analysed in detail, a genetic defect has been identified in 83% (171), as summarised in Table 6. These comprise
153 patients (74.2% of all FH patients) with 88 different point
mutations and 13 different deletions or rearrangements in
LDLR (i.e. major rearrangements comprise 13.6% of heterozygous LDLR mutations), 4 patients with two different
mutations in PCSK9 (0.2% of all FH patients) [2] and 14
FDB patients (6.8% of all FH patients) [30]. In addition we
have characterised 28 “homozygous” FH patients, 4 of whom
have recessive hypercholesterolaemia due to mutations in
LDLRAP1 (formerly known as ARH) [11,26], 13 are compound heterozygous FH patients of total or partial UK origin
who have two different mutant LDLR alleles, one compound
heterozygous FH patient of Asian Indian origin and 14 of
non-UK origin who are all true homozygous FH with two
identical LDLR alleles. All but two of the mutations found
in the homozygous FH patients of UK origin have also been
Please cite this article in press as: Tosi I et al., Genetic defects causing familial hypercholesterolaemia: Identification of deletions and
duplications in the LDL-receptor gene and summary of all mutations found in patients attending the Hammersmith Hospital Lipid Clinic,
Atherosclerosis (2006), doi:10.1016/j.atherosclerosis.2006.10.003
ATH-9692;
No. of Pages 10
8
ARTICLE IN PRESS
I. Tosi et al. / Atherosclerosis xxx (2006) xxx–xxx
Table 6
Summary of genetic defects in FH patients attending Hammersmith Lipid Clinic
Gene
Type of mutation
mutationsb
No. of different
mutations
No. of unrelated HHa probands
References
Htz
“Hmz”
153
24
[14–22] and unpublished
data
LDLR
Point
Rearrangementsc
85
13
PCSK9
Point mutationsb
2
4
0
[2]
mutationsb
4
1
0
4
[11]
ARH/LDLRAP1
Point
Rearrangementsd
APOB
None
a
b
c
d
ApoB3500
Definite FH
Possible FH
14
[30]
18
17
HH, Hammersmith Hospital; htz, heterozygous; “hmz”, homozygous or compound heterozygous.
Detected by sequencing of mRNA and/or DNA; details in Table 1.
Detected by MLPA or Southern blotting.
Detected by FISH.
found in unrelated heterozygous patients, but many of the
mutations in homozygous FH patients of non-UK origin are
unique in this study. In addition, we have identified mutations
in more than 200 affected relatives of 80 FH probands (data
not shown) (Table 6).
Analysis of DNA by MLPA was found to be a simple, rapid
and robust means of detecting major rearrangements in the
LDLR that compared favourably with other methods such as
Southern blotting and long PCR, and some large deletions can
only be detected by this method. In our experience, the quality
of the DNA is critical for successful MLPA analysis and DNA
prepared from blood by “quick” methods may not produce
reliable results. Apparent deletions of single exons may be
caused by polymorphisms or mutations involving one or a
few base pairs in the exon, as we demonstrate here with the
apparent deletion of exon 8. Therefore, for deletions of single
exons the nucleotide sequence of the exon should always
be determined, and the presence of a deletion confirmed, if
possible, by PCR across the deletion joint. In the majority of
cases in our study, the presence of a deletion or duplication
in genomic DNA was confirmed by PCR across the deletion
joint; the three exceptions were deletions of exon 1, 1–18 and
2–18, where it was impossible to predict suitable PCR primer
sequences because there are no adjacent exons 5 to exon 1
or 3 to exon 18 known to be present, and the nearest flanking
control genes were considered to be too distant. FISH can
be used to confirm the presence of large deletions, but this
requires access to fresh cells or an immortalised cell line [11].
In this study we have demonstrated the value of genetic
diagnosis of FH in two critical cases where the clinical diagnosis was unclear. Of particular interest was the detection
of deletion of exons 16 and 17, as the proband was a child
who had had very severe hypercholesterolaemia and cutaneous xanthomatas from early infancy and was suspected
of having homozygous FH. However, the diagnosis was
unclear because the mother was reported as being normocholesterolaemic, although the father was reported as having
hypercholesterolaemia caused by an LDLR mutation. A point
mutation in exon 3 (C68Y), inherited from his father, had been
identified elsewhere [31], but no further search for mutations
had been performed. The patient’s father was subsequently
referred to the Hammersmith Lipid Clinic where we carried
out further analysis of the LDLR in several family members. Although the mother had been described as having
normal plasma cholesterol levels (total plasma cholesterol
6.5 mmol/l), she was also found to carry the LDLR allele
with the deletion; her relatively normal plasma cholesterol
level can probably be explained by her diet, which results in
her being extremely lean (body mass index <18 kg/m2 ). Thus
we were able to confirm that the child was clearly compound
heterozygous FH, having inherited a C68Y allele from the
father and an allele with deletion of exons 16–17 from the
mother, and should be treated accordingly.
In the second case, the proband had been given a diagnosis
of definite heterozygous FH on the basis of a raised plasma
cholesterol value in the patient and a strong family history,
including siblings with tendon xanthomas, hypercholesterolaemia and premature death due to coronary disease. We were
able to confirm that although the affected sibling did indeed
carry a mutation in LDLR, a deletion of exons 7 and 8, the
proband did not carry this mutation. Although the proband
had hypercholesterolaemia, it was less severe than that of
the affected sibling (untreated plasma cholesterol 8.5 mmol/l
versus 13.7 mmol/l) and was probably secondary to obesity
and/or other environmental and genetic factors.
Of the 35 patients in whom we failed to detect any mutations by MLPA analysis, we had also failed to detect an
LDLR mutation by exon-by-exon sequencing of amplified
genomic DNA. Complete sequencing of the exons and flanking sequences of PCSK9 failed to reveal any sequence variants that have not also been found in non-FH individuals
[29], and no sequence variants were detected when regions
of APOB encoding LDL-receptor binding domains [12] were
sequenced in these patients. Of the patients with no known
defect, almost half had a diagnosis of definite FH, implicating
some inherited component to their disorder, although it is pos-
Please cite this article in press as: Tosi I et al., Genetic defects causing familial hypercholesterolaemia: Identification of deletions and
duplications in the LDL-receptor gene and summary of all mutations found in patients attending the Hammersmith Hospital Lipid Clinic,
Atherosclerosis (2006), doi:10.1016/j.atherosclerosis.2006.10.003
ATH-9692;
No. of Pages 10
ARTICLE IN PRESS
I. Tosi et al. / Atherosclerosis xxx (2006) xxx–xxx
sible that this may not be monogenic or may result from more
complex interactions between gene variants and the environment. Nonetheless, we conclude that despite the recent
addition of two new genes, namely LDLRAP1 and PCSK9,
to the list of genes (LDLR and APOB) in which defects are
known to be associated with severe inherited hypercholesterolaemia, there still appear to be a substantial number of
patients with a clinical diagnosis of possible or even definite
FH who have an unknown genetic defect.
Acknowledgements
Some of the patients in this study had previously been
under the care of Professor Gilbert Thompson, or of Dr.
Mary Seed at the Charing Cross Hospital, and we are
indebted to them for their participation. Bruce Pottinger
provided excellent technical assistance and MSc project
students Mafalda Bourbon and Daniel Meechan identified
some of the point mutations not previously published. Four
of our index patients with LDLR rearrangements are part
of the Simon Broome Register Study, and we are grateful to Prof. S. Humphries for sharing preliminary results.
We are grateful to colleagues who sent us samples and
entrusted to us the genetic diagnosis of their homozygous FH
patients.
References
[1] Goldstein J, Hobbs H, Brown M. Familial hypercholesterolemia. In:
Valle D, Scriver CR, Beaudet A, Sly WS, Childs B, Kinzler KW,
Volgestein B, editors. The metabolic and molecular bases of inherited
disease. New York: McGraw Hill; 2001. p. 2863–913.
[2] Sun XM, Eden ER, Tosi I, et al. Evidence for effect of mutant PCSK9 on
apolipoprotein B secretion as the cause of unusually severe dominant
hypercholesterolaemia. Hum Mol Genet 2005;14:1161–9.
[3] Top B, Koeleman BP, Gevers Leuven JA, et al. Rearrangements
in the LDL receptor gene in Dutch familial hypercholesterolemic
patients and the presence of a common 4 kb deletion. Atherosclerosis 1990;83:127–36.
[4] Lelli N, Ghisellini M, Gualdi R, et al. Characterization of three mutations of the low density lipoprotein receptor gene in Italian patients
with familial hypercholesterolemia. Arterioscler Thromb 1991;11:
234–43.
[5] Aalto-Setala K, Koivisto UM, Miettinen TA, et al. Prevalence and geographical distribution of major LDL receptor gene rearrangements in
Finland. J Intern Med 1992;231:227–34.
[6] Sun XM, Webb JC, Gudnason V, et al. Characterization of deletions in the LDL receptor gene in patients with familial hypercholesterolemia in the United Kingdom. Arterioscler Thromb 1992;12:
762–70.
[7] Garcia-Garcia AB, Real JT, Puig O, et al. Molecular genetics of familial
hypercholesterolemia in Spain: ten novel LDLR mutations and population analysis. Hum Mutat 2001;18:458–9.
[8] Schouten JP, McElgunn CJ, Waaijer R, et al. Relative quantification
of 40 nucleic acid sequences by multiplex ligation-dependent probe
amplification. Nucleic Acids Res 2002;30:e57.
[9] Scientific Steering Committee on behalf of the Simon Broome Register
Group. Risk of fatal coronary heart disease in familial hypercholesterolaemia. BMJ 1991;303:893–6.
9
[10] Sun XM, Patel DD, Knight BL, et al. Comparison of the genetic defect
with LDL-receptor activity in cultured cells from patients with a clinical
diagnosis of heterozygous familial hypercholesterolemia. The Familial
Hypercholesterolaemia Regression Study Group. Arterioscler Thromb
Vasc Biol 1997;17:3092–101.
[11] Eden ER, Patel DD, Sun XM, et al. Restoration of LDL receptor function in cells from patients with autosomal recessive hypercholesterolemia by retroviral expression of ARH1. J Clin Invest
2002;110:1695–702.
[12] Fouchier SW, Kastelein JJ, Defesche JC. Update of the molecular
basis of familial hypercholesterolemia in The Netherlands. Hum Mutat
2005;26:550–6.
[13] MRC-Holland (http://www.mrc-holland.com).
[14] Soutar AK, Knight BL, Patel DD. Identification of a point mutation in
growth factor repeat C of the low density lipoprotein-receptor gene in
a patient with homozygous familial hypercholesterolemia that affects
ligand binding and intracellular movement of receptors. Proc Natl Acad
Sci USA 1989;86:4166–70.
[15] Webb JC, Sun XM, Patel DD, et al. Characterization of two new point
mutations in the low density lipoprotein receptor genes of an English
patient with homozygous familial hypercholesterolemia. J Lipid Res
1992;33:689–98.
[16] Feher MD, Webb JC, Patel DD, et al. Cholesterol-lowering drug therapy
in a patient with receptor-negative homozygous familial hypercholesterolaemia. Atherosclerosis 1993;103:171–80.
[17] Gudnason V, King-Underwood L, Seed M, et al. Identification of
recurrent and novel mutations in exon 4 of the LDL receptor gene
in patients with familial hypercholesterolemia in the United Kingdom.
Arterioscler Thromb 1993;13:56–63.
[18] Sun XM, Patel DD, Bhatnagar D, et al. Characterization of a splicesite mutation in the gene for the LDL receptor associated with an
unpredictably severe clinical phenotype in English patients with heterozygous FH. Arterioscler Thromb Vasc Biol 1995;15:219–27.
[19] Webb JC, Patel DD, Shoulders CC, et al. Genetic variation at a splicing
branch point in intron 9 of the low density lipoprotein (LDL)-receptor
gene: a rare mutation that disrupts mRNA splicing in a patient with
familial hypercholesterolaemia and a common polymorphism. Hum
Mol Genet 1996;5:1325–31.
[20] Webb JC, Sun XM, McCarthy SN, et al. Characterization of mutations in the low density lipoprotein (LDL)-receptor gene in patients
with homozygous familial hypercholesterolemia, and frequency of
these mutations in FH patients in the United Kingdom. J Lipid Res
1996;37:368–81.
[21] Bourbon M, Fowler AM, Sun XM, et al. Inheritance of two different
alleles of the low-density lipoprotein (LDL)-receptor gene carrying the
recurrent Pro664Leu mutation in a patient with homozygous familial
hypercholesterolaemia. Clin Genet 1999;56:225–31.
[22] Naoumova RP, Neuwirth C, Pottinger B, et al. Genetic diagnosis of
familial hypercholesterolaemia: a mutation and a rare non-pathogenic
amino acid variant in the same family. Atherosclerosis 2004;174:
67–71.
[23] Sun XM, Patel DD, Webb JC, et al. Familial hypercholesterolemia
in China. Identification of mutations in the LDL-receptor gene that
result in a receptor-negative phenotype. Arterioscler Thromb 1994;14:
85–94.
[24] Pimstone SN, Sun XM, du Souich C, et al. Phenotypic variation in
heterozygous familial hypercholesterolemia: a comparison of Chinese patients with the same or similar mutations in the LDL receptor
gene in China or Canada. Arterioscler Thromb Vasc Biol 1998;18:
309–15.
[25] Mavroidis N, Traeger-Synodinos J, Kanavakis E, et al. A high incidence of mutations in exon 6 of the low-density lipoprotein receptor
gene in Greek familial hypercholesterolemia patients, including a novel
mutation. Hum Mutat 1997;9:274–6.
[26] Soutar AK, Naoumova RP, Traub LM. Genetics, clinical phenotype, and
molecular cell biology of autosomal recessive hypercholesterolemia.
Arterioscler Thromb Vasc Biol 2003;23:1963–70.
Please cite this article in press as: Tosi I et al., Genetic defects causing familial hypercholesterolaemia: Identification of deletions and
duplications in the LDL-receptor gene and summary of all mutations found in patients attending the Hammersmith Hospital Lipid Clinic,
Atherosclerosis (2006), doi:10.1016/j.atherosclerosis.2006.10.003
ATH-9692;
10
No. of Pages 10
ARTICLE IN PRESS
I. Tosi et al. / Atherosclerosis xxx (2006) xxx–xxx
[27] Gudnason V, Muller DP, Lloyd JK, et al. Response to drugs and diet
in a compound heterozygote for familial hypercholesterolaemia. Arch
Dis Child 1995;73:538–40.
[28] Navajas M, Laurent AM, Moreel JF, et al. Detection by denaturing gradient gel electrophoresis of a new polymorphism in the apolipoprotein
B gene. Hum Genet 1990;86:91–3.
[29] Kotowski IK, Pertsemlidis A, Luke A, et al. A spectrum of PCSK9 alleles contributes to plasma levels of low-density lipoprotein cholesterol.
Am J Hum Genet 2006;78:410–22.
[30] Myant NB. Familial defective apolipoprotein B-100: a review,
including some comparisons with familial hypercholesterolaemia.
Atherosclerosis 1993;104:1–18 [published erratum appears in
Atherosclerosis 1994;105(2):253].
[31] Heath KE, Humphries SE, Middleton-Price H, et al. A molecular genetic service for diagnosing individuals with familial hypercholesterolaemia (FH) in the United Kingdom. Eur J Hum Genet
2001;9:244–52.
Please cite this article in press as: Tosi I et al., Genetic defects causing familial hypercholesterolaemia: Identification of deletions and
duplications in the LDL-receptor gene and summary of all mutations found in patients attending the Hammersmith Hospital Lipid Clinic,
Atherosclerosis (2006), doi:10.1016/j.atherosclerosis.2006.10.003