© 2001 Oxford University Press
Human Molecular Genetics, 2001, Vol. 10, No. 8
815–824
Single nucleotide polymorphisms in exons of the
apo(a) kringles IV types 6 to 10 domain affect Lp(a)
plasma concentrations and have different patterns in
Africans and Caucasians
Miroslava Ogorelkova1, Hans G. Kraft1, Christian Ehnholm2 and Gerd Utermann1,+
1Institute
of Medical Biology and Human Genetics, University of Innsbruck, Schoepfstrasse 41,
6020 Innsbruck, Austria and 2National Public Health Institute, SF 00280 Helsinki, Finland
Received 7 December 2000; Revised and Accepted 12 February 2001
DDBJ/EMBL/GenBank accession nos AF158655–AF158663,
AF139730, AF139731
Lipoprotein(a) [Lp(a)] is a complex of apolipoprotein(a) [apo(a)] and low-density lipoprotein which is
associated with atherothrombotic disease. Lp(a)
plasma levels are controlled to a large extent by the
apo(a) gene locus. Known polymorphisms in the
apo(a) gene, including the kringle (K) IV-2 variable
number of tandem repeats, explain only part of the
large interindividual variability and do not explain the
differences in Lp(a) concentrations between major
human ethnic groups. Here we performed screening
for single nucleotide polymorphisms (SNPs) in exons
and flanking intron sequences of the apo(a) K IV types
6, 8, 9 and 10 which represent 1.3 kb of coding
sequence in two African (Khoi San, Black South
Africans) and one Caucasian (Tyroleans) populations
and investigated whether they affect Lp(a) levels.
Together, 768 alleles were analyzed. We identified
14 SNPs, including 11 non-synonymous SNPs (eight
of which involved conserved residues), one splice
site and two synonymous base changes. No
sequence variants common to Africans and Caucasians
were found. Several of the newly identified SNPs
showed significant effects on Lp(a) plasma concentrations. The substitutions S37F in K IV-6 and G17R
in K IV-8 were associated with Lp(a) levels significantly
below average in Africans. In contrast, the R18W
substitution in K IV-9, which occurred with a
frequency of 8% in Khoi San, resulted in a significantly
increased Lp(a) concentration. Together, our data
suggest that several SNPs in the coding sequence of
apo(a) affect Lp(a) levels. This indicates that many
SNPs may have subtle effects on the gene product.
high degree of homology with plasminogen K IV (1). One of
these (K IV-2) occurs in multiple tandemly repeated copies
which vary in number in apo(a). Apo(a) circulates in plasma
covalently bound to LDL, forming the lipoprotein(a) [Lp(a)]
particle which has atherosclerotic and prothrombotic properties.
High levels of Lp(a) are associated with myocardial infarction,
stroke, peripheral vascular disease and childhood thromboembolism (reviewed in 2).
Lp(a) levels show a large variability within a population with
>1000-fold differences between individuals. Average Lp(a)
concentrations are higher in Africans than in non-Africans and
their distributions differ markedly between populations e.g.
from a Gaussian distribution in Khoi San to distributions
highly skewed toward lower concentrations in Caucasians and
most Asians (3–6).
Twin and sib-pair analyses have revealed that the variation
in Lp(a) concentrations is under almost complete genetic
control and largely determined by variation at the apo(a) gene
locus (7–11). Three polymorphisms in the apo(a) gene with
causal effects on Lp(a) levels have been identified so far. The
number of K IV-2 repeats, which is reflected in the size of the
protein, is inversely correlated to Lp(a) plasma concentrations
in all studied populations (3,4,6,12, reviewed in 13). Cell
studies have shown that a lower secretion rate of the larger
apo(a) isoforms from hepatocytes underlies the observed effect
(14–17). However, depending on the population sample and
the type of analysis, the K IV-2 VNTR explains only from
30 to 70% of the variability of Lp(a) concentrations in a population (13). For alleles with identical numbers of K IV repeats,
the Lp(a) level can differ by a factor of 200 (18). Remarkably,
average Lp(a) concentrations associated with apo(a) alleles
with identical K IV-2 repeat number are higher in Africans
than in Caucasians (3,4). Taken together, this suggests that
sequence variation contributes to Lp(a) level variability. To
date, two SNPs with clearly established causal effects on Lp(a)
concentrations have been identified in humans. One is the +93
C→T substitution in the upstream untranslated region of
apo(a) which introduces an alternative start codon (19) but
may also affect transcription (20). This polymorphism was
found to have an impact on Lp(a) levels in Africans but not in
INTRODUCTION
Human apolipoprotein(a) [apo(a)] is a glycoprotein containing
10 different types of kringle (K) IV domains which share a
+To
whom correspondence should be addressed. Tel: +43 512 507 3450; Fax: +43 512 507 2861; Email: [email protected]
816
Human Molecular Genetics, 2001, Vol. 10, No. 8
Caucasians, where the effect was masked by linkage disequilibrium (21). The other is a donor splice site mutation in K IV-8
which results in a null allele (22).
Neither of these two polymorphisms either alone or in
combination can explain the enormous Lp(a) concentration
variability between individuals and among populations. Therefore, in addition to these polymorphisms, other sequence variations of the apo(a) gene which might differ between ethnic
groups are expected to affect Lp(a) plasma levels. This
scenario is strongly supported by the presence of polymorphisms in the non-coding regions of the gene e.g. a PNRP and
SNPs in the 5′ flanking region and the DraIII restriction polymorphism in K IV-2 introns, which reportedly are associated
with Lp(a) concentration, probably through allelic associations
with unidentified mutations with causal effects (23–25). A
factor supposed to influence Lp(a) plasma levels is the efficiency of
Lp(a) assembly.
Therefore, a candidate region for sequence variants related
to the Lp(a) concentration is the domain of the apo(a) gene
essential for Lp(a) particle formation. Lp(a) assembly requires
two steps of interaction between apo(a) and apolipoprotein B-100
(apoB-100) of low-density lipoprotein (LDL) (reviewed in 26).
As shown by in vitro studies with recombinant apo(a) deletion
mutants, in the first step apo(a) binds non-covalently to lysinerich structures in apoB-100 via the unique K IV types 6–8 (27–29).
The non-covalent association between both macromolecules is
followed by a single disulphide bridge formed between Cys45
in K IV-9 of apo(a) and Cys4326 of apoB-100 (30,31). With this
in mind we performed a systematic SNP analysis of exons
from K IV types 5, 6, 8, 9 and 10 in African (South African
Blacks, Khoi San) and Caucasian (Austrians) populations. Of
particular interest was K IV-10, because the fibrin(ogen)
binding properties of K IV-10 are thought to mediate the pathogenicity of Lp(a) (32–34). Sequence variation in this kringle
might be related to the heterogeneity in the lysine/fibrin
binding affinity observed for different Lp(a) subtypes (35–37).
A mutation in the lys binding site (W72R) has been described
in apo(a) with reduced fibrin binding from two individuals
(38). Arginine is present in this position in the chimpanzee
which has an apo(a) K IV-10 devoid of lysine/fibrin binding
capacity (39). Moreover, when mice were made transgenic for
a form of apo(a) containing Arg in this position of K IV-10 the
animals became resistant to atherosclerosis (33). All these findings implicate the K IV 6–10 domain of apo(a) as functionally
important. Some of the newly identified SNPs in the apo(a) K IV
domain reported here have a significant influence on Lp(a)
plasma concentrations suggesting that they are of physiological significance.
RESULTS
Characterization of the intron sequences flanking the
exons of the unique apo(a) kringles
The coding sequences of apo(a) kringles IV are characterized
by their high degree of homology to each other and to the K IV
of plasminogen (1). Partial intron sequences are presently only
available for K IV types 1–4 (40). Therefore, to find unique
annealing sites for specific amplification of both exons of K IV
types 6, 8, 9 and 10 and to ensure detection of potential mutations over the whole coding sequences and splice sites, we
partially sequenced the introns of K IV types 5–10 and used the
intron sequences for primer design. (The sequencing data are
available from GenBank.) The splice site sequences of the
introns of the apo(a) unique kringles are of the GT-AG type
and are analogous to those of plasminogen K IV (41), as shown
in Figure 1. Despite the high sequence homology between the
K IV exons, our sequencing analysis revealed that only those
parts of the introns which flank the exons are highly conserved
(Fig. 1).
Screening for SNPs using denaturing gradient gel
electrophoresis (DGGE)
The exons of apo(a) K IV types 6, 8, 9 and 10 were amplified
by PCR of genomic DNA (Table 1) and analyzed for mutations
using DGGE. Attachment of GC-clamps at the 5′ or 3′end of
the fragments leading to a homogenous melting profile makes
a considerable contribution to the high sensitivity of the
method (42). To avoid the predicted low melting domains
within some of the analyzed sequences after attachment of only
one GC-clamp we modified the method by including GC-clamps
of different sizes at both ends of the fragments (Fig. 1). This
resulted in homogeneous melting maps and thereby ensured
detection of potential mutations over the entire length of fragments.
In our search for SNPs in the unique apo(a) kringles we
screened three populations (Tyroleans, South African Blacks
and Khoi San). Each population sample was initially represented by more than 100 unrelated randomly selected individuals. The PCR fragments showing a variant DGGE pattern
were sequenced. As shown in Table 2, we identified two
synonymous base substitutions, nine non-synonymous base
substitutions, two known non-synonymous exchanges and a
donor splice site mutation described in detail elsewhere (22).
Two of the non-synonymous substitutions which result in
T66M in K IV-10 (43) and T23P in K IV-8 (22,44) have been
reported previously. The T23P polymorphism was designated
K IV-8 T12P by Prins et al. (44), but this is not consistent with
the numbering used by others (1). Most of the mutations had
frequencies >2% and therefore can be considered as SNPs. All
SNPs were in Hardy–Weinberg equilibrium. However, S6I and
P71T in K IV-9 and the silent mutation in K IV-10 were
detected only in single Caucasian or African individuals,
respectively. Interestingly there were no mutations common to
Caucasians and Africans, while most of the mutations identified in Khoi San, an evolutionary old human group, were also
present in South African Blacks. They were either of similar
frequency (G17R, M75T) in Khoi San and African Blacks or
frequencies of the rare allele were much higher in Khoi San
(K IV-8 R45Q; K IV-9 R18W).
Among 768 tested alleles we did not find unknown variants
resulting in amino acid changes in K IV-10. DGGE analysis of
this kringle revealed only one silent mutation in a single
Tyrolean (data not shown) and the previously described T66M
polymorphism (43). The W72R substitution found in two
white Americans and associated with a lysine binding defect
(38,45) was not observed either in the tested Caucasian population or in Africans.
Human Molecular Genetics, 2001, Vol. 10, No. 8
817
Figure 1. Sequence comparison of the exon-flanking parts of the introns of the apo(a) K IV types and K IV of plasminogen. The substituted nucleotides are indicated.
The intron sequences of plg K IV and apo(a) K IV types 1–4 are given according to Ichinose (40) and Petersen et al. (41), respectively. The exons are indicated in
brackets.
Table 1. Sequences of primers used for amplification of the exons of apo(a) kringle types 6, 8, 9 and 10
Kringle
Amplified exon
K IV-6
exon 1
Primer sequences
Annealing
temperature (°C)
Fragment sizea (bp)
F: 5′-CGGAATTTCAGCACCAACG-3′
52
180
54
181
57
235
59
267
57
234
57
251
56
177
57
200
R: 5′-(50)* AGGTTGTACCCATTTGGATAA-3′
exon 2
F: 5′-(50)* TGTAATTTGCAGTGGCCTGAC-3′
R: 5′-AGACTTCTTACCTTGTTCAGA-3′
K IV-8
exon 1
F: 5′-(50)*CTTTTCTAGGCAATACTGAGC-3′
R: 5′-(20)*GTTCCTTTTATGGCTAACATG-3′
exon 2
F: 5′-(40)*CCTTGAATATTCTCCCATC-3′
R: 5′-(20)*CCAGTATATAGATGTCTAACC-3′
K IV-9
exon 1
F: 5′-(20)* CTCATGCAATTACTGAGTATG-3′
R: 5′-(50)* GCTCCTCTTATGGTTTTAATC-3′
exon 2
F: 5′-(10)* TGTACCTGTGTTGTTTTGG-3′
R: 5′-(40)*CAATGTTCAAATGTGTAGATATC-3′
K IV-10
exon 1
F: 5′-ATTTTCAGCACCAACTGAGCAAAC-3′
R: 5′-(50)* AAAGACATACTCATTTGGGTAGTT-3′
exon 2
F: 5′-TGGAATTTCCAGTGGCCTGACA-3′
R: 5′-(50)*TCTTACCTTGTTCAGAAGGAGG-3′
The GC-clamps are indicated by an asterisk with the size (in bases) given in parentheses.
aThe size of the GC-clamps is not included.
818
Human Molecular Genetics, 2001, Vol. 10, No. 8
Table 2. SNPs in exons 1 and 2 of K IV types 6, 8, 9 and 10 of human apo(a) in Africans and Caucasians
Kringle
Exon
Base changea
Amino acid changeb
Frequency of minor allele
Tyroleans
South African Blacks
Khoi San
K IV-6
1
109 C→T
S 37 F (S3682F)
–
–
0.0272
2
111 T→C
V 91 A (V3736A)
–
–
0.0345
K IV-8
1
48 G→A
G 17 R (G3882R)
–
0.026
0.0286
66 A→C
(T3888P)c
K IV-9
1
2
K IV-10
2
0.1743
–
–
133 G→A
T 23 P
R 45 Q (R3910Q)
–
0.0327
0.1107
154 C→T
P 52 L (P3917L)
0.0089
–
–
+ 1G→A
5′splice sited
0.0531
–
–
16 G→T
S 6 I (S3985I)
0.0049
–
–
51 C→T
R 18 W (R3997W)
–
0.0146
0.08
99 T→C
None
–
0.0204
0.0212
149 C→A
P 71 T (P4083T)
–
0.005
0.0051
162 T→C
M 75 T (M4087T)
–
0.085
0.0663
63 T→C
T 66 M (M4168T)e
0.346
0.11
–
None
0.0049
–
–
124 T→G
aCounted
from position 1 of the corresponding exon.
in parentheses are according to McLean et al. (1).
cDescribed previously by Prins et al. (44).
dDescribed previously by Ogorelkova et al. (22).
eDescribed previously by Kraft et al. (43).
bPositions
Effect of the identified mutations on Lp(a) plasma
concentrations
Lp(a) concentrations, apo(a) isoforms (by protein typing), the
K IV-2 VNTR (by PFGE/genomic blotting) and some intragenic polymorphisms were known in the studied population
samples (3,21,23). Therefore, using two different statistical
approaches it was possible to estimate whether the newly identified SNPs in the apo(a) unique kringles affect Lp(a) levels
independent from known polymorphisms. However, formal
statistical analysis was not possible for P52L, S6I and P71T
substitutions which were all found in single subjects.
Three out of six SNPs detected in Africans were found to
contribute significantly to Lp(a) concentration variability
(Table 3). The S37F substitution in K IV-6 detected only in
Khoi San, and the K IV-8 G17R substitution present in Khoi
San and South African Blacks, were both associated with
significantly lower Lp(a) plasma levels for the less common
allele (P = 0.034 for S37F; P = 0.0033 for G17R). There was a
stepwise decrease consistent with a gene dose effect from
heterozygous to homozygous individuals compared with the
wild-type subjects. When the effect of the apo(a) K IV repeat
polymorphism was considered, the difference in the concentrations between the wild-type and mutant genotype groups
remained significant (P = 0.038 for S37F; P = 0.0044 for
G17R). This clearly demonstrates that the mutations have
effects on Lp(a) concentration which are independent of the
K IV-2 VNTR. The effect of the G17R substitution reached
significance for both tests only in the pooled African sample
but not in the individual populations (Table 3). This is most
likely explained by the low frequency of the rare allele (2.6%
in South African Blacks; 2.9% in Khoi San). In both populations
the influence of the variant on Lp(a) levels was in the same
direction and of the same magnitude. Because of this and since
Lp(a) levels were not significantly different between Khoi San
and Black Africans for any G17R genotype, we considered
pooling of the two samples justified.
The R18W substitution in K IV-9 was the only substitution
where the minor allele was associated with Lp(a) levels significantly above the average (Table 3). Compared with wild-type
individuals mean Lp(a) concentrations were higher in the
heterozygotes, and 2-fold increased in the homozygous mutation carriers (P = 0.031). Again, stratification for K IV repeats
did not change the result (P = 0.033). Therefore, the effect of
this SNP was not influenced by the size polymorphism of the
apo(a) gene either.
The K IV-8 45Q allele was also associated with increased
Lp(a) levels in Khoi San. However, this was significant only
upon direct comparison (P = 0.0067; Table 3) but not after
correction for the K IV effect (P > 0.08). In South African
Blacks, the R45Q substitution did not affect Lp(a) levels
(P = 0.35), which might be explained by the lower frequency
of the 45Q allele in Blacks (3.3%) versus Khoi San (11%).
Also in the pooled African sample an effect was only seen in
the unstratified data, but not when corrected for the K IV-2
effect. In Khoi San the R45Q SNP in K IV-8 was found to be
in strong linkage disequilibrium with the R18W SNP in K IV-9.
More than 80% of the 45Q carriers also had the R18W substitution. Interestingly Gln at this position of the kringle is present
in K IV of plasminogen and in apo(a) K IV types 6, 7, 9, and
10. Therefore, this mutation can be considered as a ‘reversal’
to the plasminogen K IV type. Alternatively, it might be the
older one despite its lower frequency. Together, this suggests
Human Molecular Genetics, 2001, Vol. 10, No. 8
819
Table 3. Association of SNPs in apolipoprotein(a) with Lp(a) plasma concentrations (mg/dl) in Africans
SNP
S37F in K IV-6
Population
Khoi San
G17R in K IV-8
Khoi San
Black Africans
Pooled Africans
Meas.
Exp.
Meas.
Exp.
Meas.
Exp.
Meas.
Exp.
Total population
30.1
30.0
30.6
30.5
29.5
30.2
30.2
30.3
N/N
31.0
30.2
31.8
30.8
30.2
29.8
31.2
30.4
N/M
14.0
29.7
8.4
22.8
16.1
36.1
11.9
28.9
M/M
–
–
2.7
29.3
–
–
2.7
29.3
Pa
Pa
= 0.034
Pb = 0.038
Pa
= 0.0047
Pb > 0.05
Pa
> 0.05
= 0.0033
Pb = 0.0171
Pb = 0.0044
Black Africans
Pooled Africans
SNP
R18W in K IV-9
R45Q in K IV-8
Population
Khoi San
Khoi San
Meas.
Exp
Meas.
Exp.
Meas.
Exp.
Meas.
Exp.
Total population
31.9
30.7
30.6
30.5
29.6
30.5
30.2
30.5
N/N
29.7
30.8
28.5
29.6
29.1
30.4
28.8
30.0
N/M
42.7
30.2
36.8
33.9
37.1
32.7
36.9
33.7
M/M
59.5
28.2
53.6
32.1
–
–
53.6
32.1
Pa = 0.0313
Pa = 0.0067
Pa > 0.34
Pa = 0.006
Pb = 0.033
Pb > 0.08
Pb > 0.35
Pb = 0.127
N, major allele; M, minor allele; Meas., measured Lp(a) concentration; Exp., expected Lp(a) concentration.
aSignificance of the difference between measured Lp(a) levels.
bSignificance of the difference between ∆Lp(a) levels [measured Lp(a)-expected Lp(a)].
Table 4. Mean Lp(a) concentrations (mg/dl) in relation to K IV-8 T23P
genotypes in Caucasians
Population
Tyroleans
Finns
Meas.
Exp.
Meas.
Exp.
Total population
17.1
16.7
13.1
13.2
T/T
14.8
12.2
10.8
11.4
T/P
23.1
25.8
21.7
19.6
P/P
9.5
27.1
21.8
24.1
Pa
Pb
> 0.06
Pa
> 0.1
= 0.0022
Pb
> 0.35
Meas., measured Lp(a) concentration; Exp., expected Lp(a) concentration.
aSignificance of the difference between measured Lp(a) levels.
bSignificance of the difference between ∆Lp(a) levels [measured Lp(a) - expected
Lp(a)].
that the R18W site in K IV-9 is the SNP which causally affects
Lp(a) levels and that the inconsistent effect of the R45Q substitution is caused by the allelic association with R18W.
The recently reported T23P substitution in K IV-8 (44) was
found in Tyroleans with a frequency of 17% (Table 2). The
direct comparison of Lp(a) levels between the wild-type and
mutant genotype groups did not show a significant difference
(P > 0.06; Table 4). When the data were stratified for the effect of
the K IV repeats, the 23P allele was associated with significantly
decreased Lp(a) levels (P = 0.0022). The lower than expected
Lp(a) levels were, however, only seen in homozygotes for 23P
but not in heterozygotes which, in contrast, had higher Lp(a)
than the wild-type. This inconsistency argues for a chance
finding.
To test if any effect of the T23P substitution on Lp(a) levels
could be seen in another Caucasian population, we included
104 unrelated Finnish individuals in the study.
This group was analyzed by a rapid restriction assay based
on the introduction of a HaeIII site by the mutation (44). In the
Finnish population the polymorphism had an allele frequency
of 11.5%. In the Finns, Lp(a) levels did not differ between
T23P genotypes whether compared directly or following stratification for the K IV-2 VNTR on the basis of the size alleles (Table 4).
Therefore, the effect observed in the Tyroleans most likely has
arisen by chance.
The very conservative change V91A in K IV-6 and the
M75T SNP polymorphism in K IV-9 found in Africans did not
show effects on Lp(a) levels (P > 0.4 for V91A; P > 0.35 for
M75T, data not shown).
A possible effect on Lp(a) concentration of the P52L change
could not be statistically analyzed since the mutation was identified in only two Tyroleans (Table 2). However, we noted that
the Lp(a) levels of both individuals were strikingly lower than
expected on the basis of the size alleles (Table 5). We also
noted that both individuals had one allele with 11 PNRs. Such
alleles are rare and are associated with very low Lp(a) in
Caucasians (23,24). However, the PNRP has no direct effect
on Lp(a) levels and it has therefore been suggested that a
causal mutation is in linkage disequilibrium with PNR11
(24,46). Both subjects with 52L and PNR11 possessed one
non-expressed (‘null’) apo(a) allele with 22 K IV repeats
820
Human Molecular Genetics, 2001, Vol. 10, No. 8
Table 5. Lp(a) levels and apo(a) polymorphisms in Tyrolean individuals heterozygous for K IV-8 P52L
Individual
Lp(a) (mg/dl)
+93C→T
K IV-2 VNTR
–1231
Meas.
Exp.
Genotype
Phenotype
A
0.9
28
22/32
–/32
C/C
PNRP
8/11
B
2.2
28
22/34
–/34
C/C
9/11
Meas., measured Lp(a) concentration; Exp., expected Lp(a) concentration.
revealed by PFGE and immunoblotting and they were also
heterozygous for the T23P SNP, suggesting that both mutations in
exon 1 of K IV-8 (66A→C and 154C→T) might be in cis. This
was confirmed by cloning of the PCR products and subsequent
sequencing of the inserts (data not shown). The allelic association between the less common sites together with the much
lower frequency of 52L indicates that the latter is more recent
and therefore it is likely that 52L is present on the apo(a) haplotype common to both individuals i.e. the allele with 22 K IV
repeats found to be non-expressed. Hence, 52L might be the
causal mutation responsible for the low Lp(a) associated with
11 PNRs. Family studies and in vitro expression of the mutant
apo(a) forms will be necessary to confirm this assumption.
Analysis of allelic association
We tested for allelic associations between the newly identified
SNPs and between these SNPs and other known polymorphisms in apo(a) by an exact test employing maximum likelihood calculations (56). It has already been mentioned above
that allelic associations exist between several of the analyzed
polymorphisms which makes identification of putative causal
mutations difficult. The substitutions of S37F and G17R were
found to be associated with Lp(a) levels below the average in
Africans, an effect that was independent from the K IV-2 polymorphism of the apo(a) gene. In Khoi San we found an allelic
association (P < 0.0083) between S37F in exon 1 of K IV-6
and G17R in exon 1 of K IV-8. However, the minor alleles 37F
and 17R are in repulsion indicating that the Lp(a) lowering
effects of these mutations are independent from each other. As
described above R45Q and R18W are also in strong linkage
disequilibrium in Khoi-San which may explain the marginal
but non significant association of 45Q with higher average
Lp(a) levels.
Another known polymorphism responsible for a causal
reduction of Lp(a) concentrations and present in this population is the +93C→T change in the 5′ untranslated region of the
apo(a) gene (21). We therefore tested whether allelic association
between the +93T site and the newly identified mutations in
the unique kringles might explain the observed decrease of the
Lp(a) levels. No allelic association of the SNPs with the
+93C→T polymorphism was detected.
Allelic associations were further observed between SNPs
and the K IV-2 repeat polymorphism e.g. in Caucasians the
K IV-8 T23P is associated with the K IV-VNTR. Within the
group of 23P mutation carriers from both Caucasian populations
(Tyroleans and Finns) the frequencies of the apo(a) K IV-2 alleles
ranging between 21 and 25 K IV repeats were significantly higher
than expected (P < 0.001 for both groups) while larger K IV-2
alleles had a lower than expected frequency (P < 0.05).
DISCUSSION
The apo(a) gene and protein contain three major domains.
(i) The protease domain, which is of unknown functional
significance; (ii) the 5.6 kb K IV-2 domain, which is characterized by extensive variation in copy number (2 to greater than
30) and hence difficult to analyze for mutations; and (iii) the
so-called unique kringle domain comprising K IV-1 and K IV
types 3–10 and K V (1).
Here we have analyzed a part of the coding region and
adjacent intronic sequences of the unique kringle domain for
SNPs. The aim was to see with which frequency they occur,
whether they are present in different frequencies in Africans
and Caucasians and whether such polymorphisms might have
an impact on Lp(a) concentrations and hence be of functional
significance. Due to the presence of apo(a) in a complex
[Lp(a)] with widely varying but genetically determined
concentrations, the apo(a) gene is an ideal model to study
whether SNPs have subtle effects on the gene product. Prins et al.
(47) recently analyzed the K IV types 5–10 domain in one
European population. These authors used primers in the K IV
exons. They identified three sequence variants (R60S, L87V
and L101V) in K IV-7. The latter two were in complete linkage
disequilibrium. Furthermore, they observed one SNP in K IV-8
(T12P according to their numbering, T23P in this paper) and a
variant (Y2F) in K IV-10. There was no indication of an effect
on Lp(a) levels for either of these. We have not analyzed K IV-7
here and did not observe the K IV-10 Y2F substitution in our
study. This may be due to the low frequency of the minor allele
or due to population differences. Both K IV Y2F and K IV-7
R60S are rare, occurring with a frequency of 1% in the Dutch
population (47). In our study, which investigates the frequencies of sequence variants and their effects on Lp(a) levels in
three human groups (Khoi San, a Black African population and
a European population) of distinct positions in the presumed
evolutionary tree of modern humans, we observed striking
differences in SNP patterns, frequencies and effects on Lp(a)
concentrations among populations.
An average of more than 200 independent alleles from each
of three populations, i.e. a total of 768 alleles were screened. In
the eight exons and flanking regions (representing 1328 bp
coding sequence) we identified a total of 14 SNPs, 13 in coding
sequences (cSNPs). One SNP which has been reported previously
(22) occurred at a splice site. The density of SNPs (1/95 bp;
nucleotide diversity θ = 14.6 × 10–4) in this part of the coding
sequence of apo(a) is in the higher range seen for many other
human genes (48,49). This may reflect the high number of
individuals from African populations and in particular Khoi
San or may be characteristic for the apo(a) gene or both. The
frequency of SNPs in this domain of apo(a) may be even
Human Molecular Genetics, 2001, Vol. 10, No. 8
higher because we used DDGE for SNP detection which does
not have a 100% mutation detection rate. We have, however,
directly sequenced exon 8 from 68 alleles (20 European, 26
African, 18 Asian, four Pacific) without detecting SNPs not
already known from the DGGE analysis (data not shown).
Frequencies of the minor allele ranged from 0.0049 to 0.17.
Some of the variants occurred only in single individuals and
some others were rare (≤5%). Several were of intermediate
frequency (5–15%) and none of the newly discovered SNPs
was frequent. Only the minor allele of the known K IV-10
T66M polymorphism has a frequency >30% in Caucasians
(43). The frequency of 66M is only 11% in South African
Blacks and the site is monomorphic in Khoi San. Only two of
the SNPs were synonymous whereas all others were nonsynonymous exchanges. Thus, there is a high frequency of
non-synonymous exchanges in this region of apo(a). The
domain analyzed contains K IV types 6–9 which interact with
apoB-100 of LDL and are essential for assembly of Lp(a). Not
unexpectedly, several of the polymorphisms had strong and
significant effects on Lp(a) concentrations which were independent from the K IV-2 variation and the +93C→T SNP. The
substitutions S37F in K IV-6 and G17R in K IV-8, although
present with relatively low frequencies in Africans, both
showed pronounced decreasing effects on Lp(a) concentrations. On the other hand, the change R18W in K IV-9, which
occurred with a frequency of 8% in Khoi San, was associated
with Lp(a) levels significantly higher than in the wild-type
(defined as the most common allele). Since most described
missense mutations are associated with defects in the gene
product, it is notable that in this case the amino acid substitution does increase rather than decrease the concentration of a
protein. With a frequency of 0.08, it is also unlikely that 18 W
is the ancestral allele.
Based on present knowledge, contact with LDL is the only
function related to the apo(a) unique K IV types 6–9 and, therefore, the observed effect of the sequence heterogeneity in this
region on Lp(a) concentrations could reflect either a modified
Lp(a) assembly efficiency or protein instability. The initial
non-covalent association between apo(a) and apoB-100 has
been shown as a critical step in the Lp(a) particle formation
(50). This process is thought to be mediated by the cooperative
action of weak lysine-binding sites (LBS) in K IV types 6, 7
and 8 which are, however, ‘masked’ in the intact Lp(a) molecule
and probably stabilize it (29,50–52). Apo(a) K IV-9, which is
predicted to lack a LBS, nevertheless has recently been demonstrated also to participate in the non-covalent contact between
apo(a) and LDL (29). The non-covalent step of the Lp(a)
assembly is inhibited by lysine and lysine-analogues (28,52)
but also by arginine, proline and phenylalanine (28,29) which
suggests that in addition to the LBS other sequences might be
involved in the non-covalent recognition and interaction with
apoB-100. The amino acid changes reported here are outside
the predicted lysine binding pockets but may be located within
such unidentified domains giving a clue for further in vitro
analysis.
Though we have carefully stratified for the affects of the
polymorphisms which are known to be causally related to
Lp(a) concentrations we can not exclude that some effects
observed are due to allelic associations with unidentified mutations in other parts of the gene.
821
Mutations in regulatory sequences have been related to Lp(a)
levels but their effects are much smaller than those observed
here and it is unknown whether these polymorphisms exist in
African populations. There exist strong linkage disequilibria at
the apo(a) locus some of which even extend over the K IV-2
repeat domain of the gene (25) and are present in a wide range
of world populations (H.G. Kraft and G. Utermann, unpublished
data). This makes any identification of SNP-Lp(a) level association difficult and requires rigorous stratification for potentially
confounding variants in apo(a). False positive associations
may arise if LD with K IV-2 repeat alleles are not considered
but also for other reasons, e.g. by chance. An example is the
‘significant’ association of the K IV-8 23P allele with Lp(a)
levels which were restricted to homozygotes in one population
and could not be replicated in another European population.
There are, however, arguments that those mutations identified here as affecting Lp(a) indeed are causal. Two features
were common to all missense mutations found to affect Lp(a)
levels. First, they occurred in highly conserved regions and
introduced amino acids not present at these positions in the
other apo(a) kringles or in K IV of plasminogen. Secondly, the
amino acid replacements with influence on Lp(a) concentration were non-conservative. Some, like K IV-8 P52L, involved
proline residues in the kringle domain and are expected to
modify folding of the kringles. Others may change the surface
polarity of critical domains, e.g. ligand binding sites. In vitro
expression of the mutants will be necessary to clarify whether
effects are indeed causal and, if so, which mechanisms
underlie the observed effects.
The Lp(a) concentration was influenced only by nonconservative amino acid replacements within the kringles,
whereas the V91A and changes in the inter-kringle regions did
not show any effects. Since the inter-kringle sequences are
highly variable between the different apo(a) K IV types, it may
be concluded that this part is less critical for the kringles
conformation and hence for their properties. However, the
SNPs in this region represent conservative substitutions which
may also explain lack of an effect.
The paucity of SNPs in K IV-10 contrasts the frequency of
SNPs in the other analyzed type IV kringles. This finding
might be incidental but could also indicate that this kringle has
functional characteristics which do not tolerate accumulation
of mutations. K IV-10 has the highest homology to K IV of
plasminogen and shares its pronounced lys-binding capacity.
Apo(a) K IV-10 is therefore believed to realize the interaction
of Lp(a) with biological substrates like fibrin(ogen) and cell
surface receptors (39,50–52).
SNP patterns within the unique kringles of apo(a) differed
markedly between populations. No SNP common to Caucasians
and Africans was identified in the present study. This indicates
that the observed sequence variants in the unique kringles
either originated after the split of Africans and non-Africans or
that the ancestors of the analyzed African populations did not
contribute to the European gene pool. Population-specific SNP
patterns have also been observed for other genes (48,49) but
the situation may be unique for the apo(a) gene.
Average Lp(a) levels are 3-fold higher in Africans than in
Caucasians. The factors which maintain high Lp(a) levels in
African populations are unknown. We noted, however, that
those mutations associated with decreased Lp(a) levels in Africans
S37F and G17R have low frequencies. In contrast, a splice site
822
Human Molecular Genetics, 2001, Vol. 10, No. 8
mutation in K IV-8 (Table 2) resulting in a Lp(a) null type was
found to occur with a much higher frequency in Caucasians
and was not present in Africans (22). Furthermore, an SNP
associated with Lp(a) concentration above the average R18W
was identified only in Africans. Although the identified SNPs
certainly cannot explain the difference in the median Lp(a)
levels between Africans and Caucasians, the results of this
study support the hypothesis that mutations in the coding
sequences of apo(a) contribute to the variability of Lp(a)
concentrations and may also contribute to differences in Lp(a)
levels across populations.
Beyond a better understanding of the apo(a)/Lp(a) system,
our results also have a more general implication. They demonstrate that SNPs may have rather subtle effects on a trait which,
however, may not be easily detected in other systems. The
Lp(a) system may therefore be a valuable model to study the
relation between gene variation, effects on functional properties of
the gene product, and complex disease (i.e. atherothrombotic
disease).
MATERIALS AND METHODS
Population samples
The Caucasian (Tyroleans n = 130) and African (Black South
Africans n = 104; Khoi San n = 150) population samples used
in this study have been described previously (3). The number
of alleles that were screened by DGGE ranged from 204 to 260
for the Tyroleans, from 182 to 208 for African Blacks and from
98 to 150 for the Khoi San population. Additionally, DNA
containing agarose plugs was prepared from EDTA blood
according to Kraft et al. (3) from 104 unrelated Finnish
individuals collected as a random subset of the population.
Lp(a) concentration, geno- and phenotypes of the apo(a) K IV-2
VNTR and other intragenic polymorphisms (e.g. the 5′-PNRP,
K IV-10 M66T, and +93C→T) were determined prior to this
study for all tested individuals as described elsewhere
(3,21,23,43).
Sequencing of the introns of the unique kringles
Fragments consisting of exons/introns of the unique apo(a)
kringles were amplified by PCR according to Mihalich et al.
(53) from a YAC clone (ID number ICRFy900C0535) which
contains the human apo(a) gene (54). The PCR products were
partially sequenced (see below) with primers used for the
corresponding PCR (53) or primers located in the exons (the
sequence of the primers is available on request).
PCR amplification and DGGE mutation analysis of the
exons of K IV types 6, 8, 9 and 10
The exons of K IV types 6, 8, 9 and 10 were amplified by PCR
from genomic DNA with sets of primers shown in Table 1. For
DGGE, the melting profile of the fragments was analyzed by
the computer algorithm Melt 87 (55). According to the optimal
predicting melting behavior, GC rich sequences (GC clamps)
were included at the 5′ end of one or both primers (Table 1) to
obtain the highest sensitivity of mutation detection (42). The
PCR was carried out in 50 µl final volume containing 25 pmol
of each primer, 5 µl of the melted DNA plug, 100 µM dNTP
(Boehringer Mannheim), 1.5 U DynaZyme II (Finnzymes Oy)
and 1× incubation buffer. The reaction was performed through
35 cycles with the following steps: denaturation (94°C) for 40 s,
annealing (Table 1) for 40 s and elongation (72°C) for 40 s.
The PCR started with the initial denaturation at 94°C for 5 min
and finished with a final extension at 72°C for 7 min. The PCR
products were concentrated by lyophilization and diluted in 14 µl
DGGE loading buffer (20% Ficoll 400, 10 mM Tris–HCl pH 7.5,
1 mM EDTA, 0.25% xylene cyanol and 0.25% bromophenol
blue). Prior to the DGGE run, the probes were denatured at
98% C for 3 min and allowed to reanneal at room temperature
for 10 min. DGGE was run in 9% polyacrylamide gels for 12–16 h
at 80 V (6 V/cm) in 1× TAE buffer.
The electrophoretic equipment was from C.B.S. Scientific
Co., model #DGGE-4000. The denaturing conditions were
created by 60°C constant temperature of the electrophoresis
chamber and urea/formamide gradient in the gel as follows:
14–26% formamide and 2.45–4.55 M urea for exon 1 of K IV-6;
14–30% formamide and 2.45–5.25 M urea for exon 2 of K IV-8;
12–28% formamide and 2.1–4.9 M urea for all other fragments.
The gels were stained with ethidium bromide and analyzed
under UV light.
PCR-based restriction polymorphism analysis
PCR amplification of exon 1 of K IV-8 was performed in 20 µl
total volume as described above. HaeIII digestion of the PCR
products was performed according to Prins et al. (44). The
amplicons containing C at position 66 produced two fragments
of 108 and 127 bp which were electrophoretically separated on
a 2.5% agarose gel.
Direct sequencing of PCR products
The PCR products of interest were purified on Qiaquick
columns (Qiagen) and then served as a template for a singlestrand cycle sequencing using Big Dye Terminator Cycle
Sequencing Ready Reaction Kit (Perkin-Elmer Corp.).
Sequencing reactions were carried out in 40 cycles according
to the manufacturer’s protocol. After the reaction the probes
were purified on Sepharose columns (Perkin-Elmer Corp.) and
were analyzed on an ABI PRISM 310 Genetic Analyzer. Each
PCR product was sequenced in the forward and reverse direction
with the primers used for the corresponding PCR.
Cloning of PCR products
Exons of interest were amplified as described above and
cloned in PCR™ 2.1 vector using the Original TA Cloning Kit
(Invitrogen). For the PCR, Taq DNA Polymerase (Boehringer
Mannheim) generating poly(A) ends was used. The cloning
procedures were according to the manufacturer’s protocol. The
inserts of the identified positive clones were sequenced as
described above.
Statistical analysis
An exact test based on the maximum likelihood method was
used to generate haplotypes and to test whether the identified
polymorphisms were in Hardy–Weinberg equilibrium. The
presence of allelic associations was identified by the likelihood
ratio test of linkage disequilibrium. The phase of the allelic
association was identified using a χ2 goodness-of-fit test.
These calculations were performed using Arlequin (56).
Human Molecular Genetics, 2001, Vol. 10, No. 8
We used two approaches to test for the significance of an
SNP effect on Lp(a) levels. (i) Lp(a) levels of wild-type individuals and hetero- and homozygotes for the respective SNPs
were directly compared by the non-parametric Kruskal–Wallis
one-way analysis of variance (ANOVA). (ii) Linear multiple
regression analysis considering the effect of the K IV-2 VNTR
was applied as described (21,23) to calculate average K IV-2
allele associated Lp(a) levels [designated expected Lp(a)
levels] in each population. The effect of a mutation on Lp(a)
concentration was then estimated by the difference between
measured and expected Lp(a) level and defined as ∆Lp(a).
∆Lp(a) levels were compared between the wild-type and
mutant genotypes using Kruskal–Wallis one-way ANOVA.
These calculations were performed using the SPSS for
Windows package (release 9.0). A difference in Lp(a) levels
was only considered significant if (i) both approaches showed
significant differences between genotypes at P < 0.05 and
(ii) if there was a gene dose effect.
ACKNOWLEDGEMENTS
We thank Dr Rhena Delport (Pretoria) for the African population samples, and Susanne Rauchenwald for expert technical
assistance. This work was supported by grant P 12819-GEN
from the Austrian Science Fund to G.U.
REFERENCES
1. McLean, J.W., Tomlinson, J.E., Kuang, W.-J., Eaton, D.L., Chen, E.Y.,
Fless, G.M., Scanu, A.M. and Lawn, R.M. (1987) cDNA sequence of
human apolipoprotein (a) is homologous to plasminogen. Nature, 300,
132–137.
2. Utermann, G. (2000) Lipoprotein(a). In Scriver, Ch.R., Beaudet, A.L.,
Sly, W.S. and Valle, D. (eds), The Metabolic and Molecular Bases of
Inherited Disease. McGraw Hill Inc., New York, pp. 2753–2787.
3. Kraft, H.G., Lingenhel, A., Pang, R.W.C., Delport, R., Trommsdorff, M.,
Vermaak, H., Janus, E.D. and Utermann, G. (1996) Frequency distributions of apolipoprotein(a) kringle IV repeat alleles and their effects on
lipoprotein(a) levels in Caucasian, Asian, and African Populations: The
distribution of null alleles is non-random. Eur. J. Hum. Genet., 4, 74–87.
4. Gaw, A., Boerwinkle, E., Cohen, J.C. and Hobbs, H.H. (1994) Comparative
analysis of the apo(a) gene, apo(a) glycoprotein, and plasma concentrations of Lp(a) in three ethnic groups. Evidence for no common “null”
allele at the apo(a) locus. J. Clin. Invest., 93, 2526–2534.
5. Marcovina, S.M., Albers, J.J., Jacobs, D.R., Jr, Perkins, L.L., Lewis, C.E.,
Howard, B.V. and Savage, P. (1993) Lipoprotein[a] concentrations and
apolipoprotein[a] phenotypes in Caucasians and African Americans: The
CARDIA Study. Arterioscler. Thromb., 13, 1037–1045.
6. Sandholzer, C., Hallman, D.M., Saha, N., Sigurdsson, G., Lackner, C.,
Császár, A., Boerwinkle, E. and Utermann, G. (1991) Effects of the apolipoprotein(a) size polymorphism on the lipoprotein(a) concentration in 7
ethnic groups. Hum. Genet., 86, 607–614.
7. Kraft, H.G., Köchl, S., Menzel, H.J., Sandholzer, C. and Utermann, G.
(1992) The apolipoprotein(a) gene: a transcribed hypervariable locus controlling plasma lipoprotein(a) concentration. Hum. Genet., 90, 220–230.
8. Boerwinkle, E., Leffert, C.C., Lin, J., Lackner, C., Chiesa, G. and Hobbs,
H.H. (1992) Apolipoprotein(a) gene accounts for greater than 90% of the
variation in plasma lipoprotein(a) concentrations. J. Clin. Invest., 90, 52–60.
9. DeMeester, C.A., Bu, X., Gray, R.J., Lusis, A.J. and Rotter, J.I. (1995)
Genetic variation in lipoprotein(a) levels in families enriched for coronary
artery disease is determined almost entirely by the apolipoprotein(a) gene
locus. Am. J. Hum. Genet., 56, 287–293.
10. Mooser, V., Scheer, D., Marcovina, S.M., Wang, J., Guerra, R., Cohen, J.
and Hobbs, H.H. (1997) The apo(a) gene is the major determinant of
variation in plasma Lp(a) levels in African Americans. Am. J. Hum.
Genet., 61, 402–417.
823
11. Scholz, M., Kraft, H.G., Lingenhel, A., Delport, R., Vorster, E.H., Bickeböller,
H. and Utermann, G. (1999) Genetic control of lipoprotein(a) concentrations is
different in Africans and Caucasians. Eur. J. Hum. Genet., 7, 169–178.
12. Utermann, G., Menzel, H.J., Kraft, H.G., Duba, H.C., Kemmler, H.G. and
Seitz, C. (1987) Lp(a) glycoprotein phenotypes. Inheritance and relation
to Lp(a)-lipoprotein concentrations in plasma. J. Clin. Invest., 80, 458–465.
13. Utermann, G. (1999) Genetic architecture and evolution of the lipoprotein(a)
trait. Curr. Opin. Lipid., 10, 133–141.
14. Brunner, C., Lobentanz, E.M., Petho-Schramm, A., Ernst, A., Kang, C.,
Dieplinger, H., Muller, H.J. and Utermann, G. (1996) The number of identical kringle IV repeats in apolipoprotein(a) affects its processing and
secretion by HepG2 cells. J. Biol. Chem., 271, 32403–32410.
15. White, A.L., Hixson, J.E., Rainwater, D.L. and Lanford, R.E. (1994)
Molecular basis for “null” lipoprotein(a) phenotypes and the influence of
apolipoprotein(a) size on plasma lipoprotein(a) level in the baboon.
J. Biol. Chem., 269, 9060–9066.
16. White, A.L., Guerra, B. and Lanford, R.E. (1997) Influence of allelic variation on apolipoprotein(a) folding in the endoplasmic reticulum. J. Biol.
Chem., 272, 5048–5055.
17. Lobentanz, E.M., Krasznai, K., Gruber, A., Brunner, C., Muller, H.J., Sattler, J.,
Kraft, H.G., Utermann, G. and Dieplinger, H. (1998) Intracellular metabolism
of human apolipoprotein(a) in stably transfected Hep G2 cells. Biochemistry,
37, 5417–5425.
18. Cohen, J.C., Chiesa, G. and Hobbs, H.H. (1993) Sequence polymorphisms
in the apolipoprotein(a) gene. J. Clin. Invest., 91, 1630–1636.
19. Zysow, R.R., Lindahl, G.E., Wade, D.P., Knight, B.L. and Lawn, R.M
(1995) C/T polymorphism in the 5′ untranslated region of the apolipoprotein(a) gene introduces an upstream ATG and reduces in vitro translation.
Arterioscler. Thromb., 15, 58–64.
20. Suzuki, S., Takahashi, H., Sato, H., Takashima, N., Arakawa, M. and
Gejyo, F. (1997) Significance of glomerular deposition of apolipoprotein
(a) in various glomerulopathies. Am. J. Nephrol., 17, 499–504.
21. Kraft, H.G., Windegger, M., Menzel, H.J. and Utermann, G. (1998)
Significant impact of the +93 C/T polymorphism in the apolipoprotein(a)
gene on Lp(a) concentrations in Africans but not in Caucasians: confounding
effect of linkage disequilibrium. Hum. Mol. Genet., 7, 257–264.
22. Ogorelkova, M., Gruber, A. and Utermann, G. (1999) Molecular basis of congenital Lp(a) deficiency: a frequent apo(a) “null” mutation in Caucasians.
Hum. Mol. Genet., 8, 2087–2096.
23. Trommsdorff, M., Köchl, S., Lingenhel, A., Kronenberg, F., Delport, R.,
Vermaak, H., Lemming, L., Faergeman, O., Utermann, G. and Kraft, H.G.
(1995) A pentanucleotide repeat polymorphism in the 5′ control region of
the apolipoprotein(a) gene is associated with lipoprotein(a) plasma concentrations in Caucasians. J. Clin. Invest., 96, 150–157.
24. Mancini, F.P., Mooser, V., Guerra, R. and Hobbs, H.H. (1995) Sequence
microheterogeneity in apolipoprotein(a) gene repeats and the relationship
to plasma Lp(a) levels. Hum. Mol. Genet., 4, 1535–1542.
25. Puckey, L.H., Lawn, R.M. and Knight, B.L. (1997) Polymorphisms in the
apolipoprotein(a) gene and their relationship to allele size and plasma
lipoprotein(a) concentration. Hum. Mol. Genet., 6, 1099–1107.
26. Dieplinger, H. and Utermann, G. (1999) The seventh myth of lipoprotein(a): where and how is it assembled. Curr. Opin. Lipid., 10, 275–283.
27. Trieu, V.N. and McConathy, W.J. (1995) A two-step model for lipoprotein(a) formation. J. Biol. Chem., 270, 15471–15474.
28. Frank, S., Durovic, S., Kostner, K. and Kostner, G.M. (1995) Inhibitors
for the in vitro assembly of Lp(a). Arterioscler. Thromb. Vasc. Biol., 15,
1774–1780.
29. Gabel, B.R. and Koschinsky, M.L. (1998) Sequences within apolipoprotein(a) kringle IV types 6–8 bind directly to low-density lipoprotein
and mediate no covalent association of apolipoprotein(a) with apolipoprotein B-100. Biochemistry, 37, 7892–7898.
30. Brunner Ch., Kraft, H.G., Utermann, G. and Müller, H.J. (1993) Cys4057
of apolipoprotein(a) is essential for lipoprotein(a) assembly. Proc. Natl
Acad. Sci. USA, 90, 11643–11647.
31. McCormick, S.P., Ng, J.K., Taylor, S., Flynn, L.M., Hammer, R.E. and
Young, S.G. (1995) Mutagenesis of the human apolipoprotein B gene in a
yeast artificial chromosome reveals the site of attachment for apolipoprotein(a). Proc. Natl Acad. Sci. USA, 92, 10147–10151.
32. LoGrasso, P.V., Cornell-Kennon, S. and Boettcher, B.R. (1994) Cloning,
expression, and characterization of human apolipoprotein(a) Kringle
IV37. J. Biol. Chem., 269, 21820–21827.
33. Boonmark, N.W., Lou, X.J., Yang, Z.J., Schwartz, K., Zhang, J.L., Rubin,
E.M. and Lawn, R.M. (1997) Modification of apolipoprotein(a) lysine
824
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
Human Molecular Genetics, 2001, Vol. 10, No. 8
binding site reduces atherosclerosis in transgenic mice. J. Clin. Invest.,
100, 558–564.
Hughes, S.D., Lou, X.J., Ighani, S., Verstuyft, J., Grainger, D.J., Lawn,
R.M. and Rubin, E.M. (1997) Lipoprotein(a) vascular accumulation in
mice. In vivo analysis of the role of lysine binding sites using recombinant
adenovirus. J. Clin. Invest., 100, 1493–1500.
Hervio, L., Chapman, M.J., Thillet, J., Loyau, S. and Anglés-Cano, E.
(1993) Does apolipoprotein(a) heterogeneity influence lipoprotein(a)
effects on fibrinolysis? Blood, 82, 392–397.
Hervio, L., Girard-Globa, A., Durlach, V. and Angles-Cano, E. (1996)
The antifibrinolytic effect of lipoprotein(a) in heterozygous subjects is
modulated by the relative concentration of each of the apolipoprotein(a)
isoforms and their affinity for fibrin. Eur. J. Clin. Invest., 26, 411–417.
Hoover-Plow, J.L., Boonmark, N., Skocir, P., Lawn, R. and Plow, E.F.
(1996) A quantitative immunoassay for the lysine-binding function of
lipoprotein(a). Application to recombinant apo(a) and lipoprotein(a) in
plasma. Arterioscler. Thromb. Vasc. Biol., 16, 656–664.
Scanu, A.M., Pfaffinger, D., Lee, J.C. and Hinman, J. (1994) A single
point mutation (Trp72→Arg) in human apo(a) kringle 4–37 associated
with a lysine binding defect in Lp(a). Biochim. Biophys. Acta Mol. Basis
Dis., 1227, 41–45.
Chenivesse, X., Huby, T., Wickins, J., Chapman, J. and Thillet, J. (1998)
Molecular cloning of the cDNA encoding the carboxy-terminal domain of
chimpanzee apolipoprotein(a): an Asp57→Asn mutation in kringle IV-10
is associated with poor fibrin binding. Biochemistry, 37, 7213–7223.
Ichinose, A. (1992) Multiple members of the plasminogen–apolipoprotein(a) gene family associated with thrombosis. Biochemistry, 31, 3113–
3118.
Petersen, T.E., Martzen, M.R., Ichinose, A. and Davie, E.W. (1990)
Characterization of the gene for human plasminogen, a key proenzyme in
the fibrinolytic system. J. Biol. Chem., 265, 6104–6111.
Sheffield, V.C., Cox, D.R., Lerman, L.S. and Myers, R.M. (1989) Attachment of a 40-base-pair G + C-rich sequence (GC-clamp) to genomic DNA
fragments by the polymerase chain reaction results in improved detection
of single-base changes. Proc. Natl Acad. Sci. USA, 86, 232–236.
Kraft, H.G., Haibach, C., Lingenhel, A., Brunner, C., Trommsdorff, M.,
Kronenberg, F., Müller, H.J. and Utermann, G. (1995) Sequence polymorphism in kringle IV 37 in linkage disequilibrium with the apolipoprotein(a) size polymorphism. Hum. Genet., 95, 275–282.
Prins, J., Leus, F.R., Van der Hoek, Y.Y., Kastelein, J.J., Bouma, B.N. and
van Rijn, H.J. (1997) The identification and significance of a Thr→Pro polymorphism in kringle IV type 8 of apolipoprotein(a). Thromb. Haemost., 77,
949–954.
45. Klezovitch, O. and Scanu, A.M. (1996) Lys and fibrinogen binding of
wild-type (Trp72) and mutant (Arg72) human apo(a) kringle IV-10
expressed in E coli and CHO cells. Arterioscler. Thromb. Vasc. Biol., 16,
392–398.
46. Bopp, S., Köchl, S., Acquati, F., Magnaghi, P., Petho Schramm, A., Kraft,
H.G., Utermann, G., Muller, H.J. and Taramelli, R. (1995) Ten allelic
apolipoprotein[a] 5′ flanking fragments exhibit comparable promoter
activities in HepG2 cells. J. Lipid. Res., 36, 1721–1728.
47. Prins, J., Leus, F.R., Bouma, B.N. and vanRijn, H.J. (1999) The identification of polymorphisms in the coding region of the apolipoprotein(a)
gene-association with earlier identified polymorphic sites and influence
on the lipoprotein(a) concentration. Thromb. Haemost., 82, 1709–1717.
48. Cargill, M., Altshuler, D., Ireland, J., Sklar, P., Ardlie, K., Patil, N., Lane,
C.R., Lim, E.P., Kalyanaraman, N., Nemesh, J. et al. (1999) Characterization of single-nucleotide polymorphisms in coding regions of human
genes. Nat. Genet., 22, 231–238.
49. Halushka, M.K., Fan, J.B., Bentley, K., Hsie, L., Shen, N., Weder, A.,
Cooper, R., Lipshutz, R. and Chakravarti, A. (1999) Patterns of singlenucleotide polymorphisms in candidate genes for blood-pressure homeostasis. Nat. Genet., 22, 239–247.
50. Frank, S. and Kostner, G.M. (1997) The role of apo-(a) kringle-IVs in the
assembly of lipoprotein-(a). Protein Eng., 10, 291–298.
51. Ernst, A., Helmhold, M., Brunner, C., Pethö-Schramm, A., Armstrong,
V.W. and Müller, H.J. (1995) Identification of two functionally distinct
lysine-binding sites in kringle 37 and in kringles 32–36 of human apolipoprotein(a). J. Biol. Chem., 270, 6227–6234.
52. Gabel, B.R., May, L.F., Marcovina, S.M. and Koschinsky, M.L. (1996)
Lipoprotein(a) assembly. Quantitative assessment of the role of apo(a)
kringle IV types 2–10 in particle formation. Arterioscler. Thromb. Vasc.
Biol., 16, 1559–1567.
53. Mihalich, A., Magnaghi, P., Sessa, L., Trubia, M., Acquati, F. and
Taramelli, R. (1997) Genomic structure and organization of kringles type
3 to 10 of the apolipoprotein(a) gene in 6q26–27. Gene, 196, 1–8.
54. Haibach, C., Kraft, H.G., Kochl, S., Abe, A. and Utermann, G. (1998) The
number of kringle IV repeats 3–10 is invariable in the human apo(a) gene.
Gene, 208, 253–258.
55. Lerman, L.S. and Silverstein, K. (1987) Computational simulation of DNA
melting and its application to denaturing gradient gel electrophoresis. Methods
Enzymol., 155, 501–527.
56. Schneider, S., Rössli, D. and Excoffier, L. (2000) Arlequin: A Software for
Population Genetics Data Analysis. Version 2.000. Department of
Anthropology, University of Geneva, Geneva, Switzerland.