A functional polymorphism within plasminogen
activator urokinase (PLAU) is associated with
Matthias Riemenschneider1,*, Lidija Konta1, Patricia Friedrich1, Sandra Schwarz1,
Kevin Taddei5,13, Frauke Neff2, Alessandro Padovani6, Heike Ko ¨lsch7, Simon M. Laws1,13,
Norman Klopp8, Heike Bickebo ¨ller9, Stefan Wagenpfeil3, Jakob C. Mueller1,3,14,
Albert Rosenberger9, Janine Diehl-Schmid4, Silvana Archetti6, Nicola Lautenschlager10,
Barbara Borroni6, Ulrich Mu ¨ller11, Thomas Illig8, Reinhard Heun7,
Rupert Egensperger12, Ju ¨rgen Schlegel2, Hans Fo ¨rstl4,
Ralph N. Martins5,13the German Sib-Pair Study Group and Alexander Kurz4
1Neurochemistry and Neurogenetics Laboratory, Department of Psychiatry and Psychotherapy,2Institute of
Pathology,3Department of Medical Statistics and Epidemiology ,4Department of Psychiatry and Psychotherapy,
Technische Universita ¨t Mu ¨nchen (TUM), Ismaningerstr. 22, 81675 Munich, Germany,5Sir James McCusker
Alzheimer’s Disease Research Unit, Hollywood Private Hospital, Perth, Western Australia,6Clinica Neurologica,
Universita ` degli Studi di Brescia, Brescia, Italy,7Department of Psychiatry and Psychotherapy, Friedrich Wilhelms
Universita ¨t, Bonn, Germany,8Institute of Epidemiology, GSF, Mu ¨nchen-Neuherberg, Germany,9Department of
Genetic Epidemiology, University of Go ¨ttingen, Germany,10School of Psychiatry and Clinical Neurosciences,
University of Western Australia, Perth, Australia,11Institute of Human Genetics, University of Giessen, Germany,
12Institute of Neuropathology, University Hospital Muenster, Germany,13Alzheimer’s and Aging, School of Biomedical
and Sports Science, Edith Cowan University, Joondalup, Australia and14Hertie-Institute for Clinical Brain Research,
Received March 22, 2006; Revised and Accepted July 1, 2006
A number of susceptibility loci for Alzheimer’s disease (AD) have been identified including a region on
Chromosome 10q21–q22. Within this region the plasminogen activator urokinase gene (PLAU) was
considered as a reasonable candidate from its functional implication in plasmin generation, a serine
protease capable of degrading beta-Amyloid (Ab) protein. We screened 56 single nucleotide polymorphisms
(SNPs) around PLAU using 1751 individuals from four independent case–control samples (Munich,
N 5 679; Bonn N 5 282; Brescia (Italy) N 5 219; Perth (Australia) N 5 557 and one discordant sib-pair
sample (Munich N 5 622). In brain tissue samples of neuropathologically confirmed cases with AD
(N 5 33) we analyzed plaque counts according to the risk allele. We identified that one functional
exonic SNP (rs2227564) is associated with development of AD using the four independent case–control
samples (Munich, P 5 0.02; Bonn, P 5 0.005; Brescia (Italy), P 5 0.001; Perth (Australia), P 5 0.03) and the
discordant sib-pair sample (P 5 0.001). In brain tissue, from neuropathologically confirmed cases with AD,
we identified significantly higher plaque counts in carriers of the risk allele (N 5 6; 60.3 +
with non-carriers (N 5 9; 26.3 +++++ 8.8; P 5 0.007). This study provides compelling evidence of a genetic and
functional involvement of a common PLAU variant into the pathogenesis of AD. Further functional investi-
gations are warranted to elucidate the specific role of PLAU, respectively, PLAU variants in the metabolism
of Ab proteins.
++++ 16.9) compared
# The Author 2006. Published by Oxford University Press. All rights reserved.
For Permissions, please email: firstname.lastname@example.org
*To whom correspondence should be addressed: Tel: þ49 8941404214; Fax: þ49 8941404888; Email: email@example.com
Human Molecular Genetics, 2006, Vol. 15, No. 16
Advance Access published on July 6, 2006
Alzheimer’s disease (AD) is a common, genetically complex
neurodegenerative disorder that causes a progressive decline
of cognitive abilities. The spectrum of factors that contribute
risk to AD is wide and ranges from rare mutations causing
an autosomal-dominant early-onset form of AD, to suscepti-
bility genes and further to epigenetic and non-genetic risk
factors, such as advanced age. Abundant evidence suggests
that elevated amounts of the small, highly amyloidogenic
protein beta-amyloid (Ab), which is mainly composed of
40–42 amino acids (Ab40, Ab42) contributes to the neuro-
degenerative process in AD through its neurotoxic properties
(1). While an increased production of Ab peptides has been
demonstrated for the rare autosomal dominant forms of
AD due to mutations within the amyloid precursor protein
gene (APP; OMIM?104760) or the presenilins (PSEN1,
OMIM?104311; PSEN2, OMIM?600759), a decreased degra-
dation of Ab proteins may contribute to the sporadic forms
of the disease. Ongoing research identified several proteases
capable of degrading Ab proteins. Among these the insulin
degrading enzyme (IDE, OMIM?146680), plasmin and activa-
tors of the plasmin system are of particular interest, as they
may represent reasonable candidate genes conferring risk to
the development of AD.
Previous genome-wide linkage studies in patients with
late-onset AD have identified several regions with consider-
able linkage, (2–7) which strongly suggests the existence of
additional susceptibility genes
(APOE), the only unequivocally confirmed risk gene of AD
so far (8,9). One broad linkage region on Chromosome 10q
[10q21–q24; multipoint LOD score (MLS) ?4] has gained
increasing attention due to the linkage to this region being
repeatedly demonstrated in independent patient samples
(10,11) and in addition, by a linkage study using elevated
Ab42 levels in plasma as a quantitative trait (12). The
genetic basis of this linkage regions has yet to be determined;
however, there is increasing evidence from functional and
genetic studies to suggest that two proteases, the plasminogen
activator urokinase (PLAU, OMIM?191840) located at the
middle and insulin degrading enzyme (IDE) at the distal end
of the Chromosome 10 linkage region are strong functional
and positional candidate genes. PLAU encodes the urokinase-
type plasmin activator (uPA), a serine protease, which acti-
vates plasmin from its inactive precursor plasminogen after
binding to its receptor (uPAR, OMIM?173391). In the
context of AD-related pathological mechanisms it has been
shown that Ab induces the expression of PLAU and its recep-
tor (PLAUR) (13,14) and that uPA was found to inhibit
Ab-mediated neurotoxicity and fibrillogenesis via activation
of plasmin, a serine protease capable of degrading aggregated
and soluble forms of Ab proteins (15,16). Besides its capa-
bility to degrade Ab peptides directly, plasmin has also been
shown to enhance the non-amyloidogenic processing of APP
by a-secretase cleavage (17).
Thus, several groups have investigated genetic variations in
PLAU for association with AD (18–21) and AD-related endo-
phenotypes, such as plasma Ab42 levels (22), cerebrospinal
fluid (CSF) Ab42 levels and cognitive performance as
assessed by the Mini Mental State Examination (MMSE)
(23). However, the results of these genetic studies remain
controversial, with some studies unable to find any association
with AD (19–21) and others, which demonstrated significant
association with AD (18,22), plasma Ab42 concentrations
(22), CSF Ab42 concentrations and MMSE (23). The situation
of PLAU appears to be even more puzzling, as Myers et al.
(19) excluded PLAU as being the major locus underlying
their linkage signal, whereas Ertekin-Taner (22) identified a
substantial contribution of PLAU to their QTL-plasma Ab42
signal and Finckh et al. (18) published a risk association to
the opposite allele compared with Ertekin-Taner et al. (22).
In an attempt to clarify the discrepancy of association
results, we applied a whole gene approach (24) and conducted
a detailed analysis of the PLAU gene locus by genotyping 56
single nucleotide polymorphisms (SNPs) spanning a genomic
region of ?300 kB around the PLAU gene using a large
case–control sample from Germany. The most promising
association signal of a coding polymorphism at Exon 6 was
successfully replicated using three independent case–control
samples and one discordant sib-pair sample. Another coding
polymorphism was identified by sequencing, but excluded as
a risk variant through genotyping. Finally, possible functional
consequences on the cerebral plaque load were investigated
using human brain samples with AD.
The initial genetic and statistical analyses to determine the LD
structure, the single marker and haplotype association signals,
as well as the sliding window approach were performed using
the Munich cohort.
Within the Munich sample, the LD structure based on 52 out
of 53 SNPs (one SNP, rs2227579, was excluded because of
deviation from Hardy–Weinberg equilibrium; Fig. 1A),
shows that the PLAU region is partially separated by likely
CAMK2G upstream (between markers hCV27069354 and
rs2675675) and completely unlinked to the VCL downstream
(between markers rs2688624 and rs4745730). Continuous
blocks of relatively high r2values are observed at the 50flank-
ing and promoter region of PLAU (from rs2675675 to
rs2227552), and from PLAU Exon 11 (rs4065) to the
30-UTR region (rs2688624). Accordingly, the markers at the
PLAU gene form two blocks of high D0values (Block 3 and
Block 4) separated by a partial breakpoint located at and
including the marker rs2227564 (PLAU Exon 6; Fig. 1B). In
addition, there were some scattered long-distance correlations
between several markers within the genotyped region between
rs2292307 and rs2688624, but these were uncorrelated with
the markers at the VCL gene.
Association with AD
Analysis of PLAU SNP alleles in the screening sample from
Munich revealed evidence of an allelic association with AD
for several markers located at the intergenic region upstream
Human Molecular Genetics, 2006, Vol. 15, No. 162447
of PLAU, the LD breakpoint region within PLAU around Exon
6, and the 30-UTR region. Specifically, three genetic markers
within PLAU (rs2227562 Intron 5, rs2227564 Exon 6,
rs2227568 Exon 8) were found to be associated with AD
(Table 1). The risk conferred to AD was associated with the
minor T allele for marker rs2227564, but was due to the
major allele for both neighboring markers rs2227562 and
rs2227568. The latter association pattern may be described
by a common risk variant with a frequency of ?86% or, alter-
natively, by a rare protective variant (frequency ?14%). Both
markers were also highly intercorrelated (r2¼ 0.9; D0¼ 0.96),
whereas the marker rs2227564 was separated from this corre-
lated cluster (pairwise comparison with rs2227562: D0¼ 0.73,
r2¼ 0.036 and rs2227568: D0¼ 0.63, r2¼ 0.025).
A more detailed analysis of the association signals within
the Munich screening sample revealed differences in the
strength of the association according to the disease onset. To
further explore this association we separately investigated
patients with an early-onset of the disease (EOAD; onset
?65 years; N ¼ 140) and those with late-onset AD (LOAD
.65 years; N ¼ 282). In the EOAD group, we observed
stronger allelic association (P ¼ 0.001) and a risk allele
(T-allele) dose-dependent increase of the odds ratio (OR)
from 1.7 (1.1–2.6) in heterozygous subjects up to 3.0
(1.1–8.0) in homozygous subjects (Table 2). No significant
association was obtained in patients with LOAD (Table 1).
Concerning both neighboring SNPs, the opposite effect was
observed. Both markers show a trend or significant allelic
association with LOAD [rs2227562: P ¼ 0.04; OR ¼ 1.5
(1.01–2.2) and rs2227568: P ¼ 0.05; OR ¼ 1.47 (0.99–2.1)],
whereas no association was present in patients with EOAD
(Table 1). Because of the number of SNPs genotyped at this
region the genotypic association signals would no longer be sig-
Two LD blocks which include the associated markers
rs2227562 (Block 3; Fig. 1B) and rs2227568 (Block 4;
Fig. 1B) were identified at the PLAU region. Within Block 3,
three haplotypes are estimated. The most common haplotype
shows a non-significant trend for association with AD
(x2¼ 3.72; P ¼ 0.054; Table 3) and another haplotype with a
nificant association (x2¼ 5.7; P ¼ 0.017; Table 3). Within
Figure 1. (A) Linkage disequilibrium structure of the extended PLAU gene region including the neighboring genes. (B) Linkage disequilibrium structure of the
PLAU gene region. Linkage disequilibrium structure of the PLAU gene region. Pairwise r2values are color-coded: black, high r2values; white, low r2values.
The haplotype blocks indicated by high D0values are superimposed. The gene structure of the genomic region is shown on top.
2448 Human Molecular Genetics, 2006, Vol. 15, No. 16
Block 4, four haplotpyes are estimated and one haplotype
showing a frequency of 12% in cases and 16.8% in controls
was significantly associated (x2¼ 6.1; P ¼ 0.014; Table 3).
Using a sliding window approach for three marker associ-
ations, we identified two regions showing evidence for a table-
wide significant association. The proximal signal was obtained
at the intergenic region within window 1 (Haplotype 1.3;
adjusted P ¼ 0.024) and a trend for window 2 (Haplotype 2.3;
adjusted P ¼ 0.051; Supplementary Table, S1). The strongest,
second signal was observed at the breakpoint for window 11
(marker combinations rs2227552, rs2227562, rs2227564) and
window 12 (marker combination rs2227562, rs2227564,
rs2227568; Supplementary Table, S1). Two haplotypes with
11.3; adjusted P ¼ 0.057) and 0.7% in cases and 3% in controls
(Haplotype 11.4; adjusted P ¼ 0.009) showed a trend or table-
wide significant association (1000 permutations). In addition,
of 11.3% in cases and 16.7% in controls was also significantly
associated (adjusted P ¼ 0.048).
To identify additional SNPs which are in LD with the func-
tional rs2227564 polymorphism, we sequenced the complete
Exon 6 of the PLAU gene including both exon/intron bound-
aries using an independent sample from Munich consisting of
80 patients with AD and 80 controls. By sequencing we ident-
ified one additional functional, but rare variation within Exon
6 (1811 C/T His ! Tyr), 23 bp upstream of the rs2227564
(1788 C/T; Leu ! Pro) polymorphism (Fig. 2). Subsequent
genotyping of this novel polymorphism in the Munich
sample did not reveal any evidence of an association with
AD or patients with EOAD (data not shown).
To replicate our initial genetic results we investigated the
rs2227564 polymorphism in three independent case–control
series from Bonn (Germany), Brescia (Italy), Perth (Australia)
and in addition applied a family-based approach using a dis-
cordant sib-pair sample from Germany.
Figure 1. Continued.
Human Molecular Genetics, 2006, Vol. 15, No. 162449
Regarding the case–control collectives we observed a sig-
nificant allelic association for all series and further obtained
a T-allele dose-dependent increase of the OR in the sample
from Bonn and a similar trend in the series from Brescia,
which just failed to reach significance in homozygous carriers
of the T allele, probably because of the limited sample size
(Table 2). In the Australian series, this effect was, likewise
to the Munich sample, only present in patients with a
younger onset of the disease. Because of a relatively small
number of patients with EOAD, we divided this sample into
two equal-sized subgroups defined by the cohort median age
of 80 years (Table 2).
As a final step of our replication procedure, we applied a
family-based strategy and analyzed the PLAU rs2227564 poly-
morphism in a discordant sib-pair sample from Germany. For
the statistical analysis we applied the S-TDT (25) and obtained
a highly significant over-representation of the risk-allele in the
affected group (P ¼ 0.001; Table 2).
rs2227564 polymorphism and the APOEe4 carrier status was
observed in the larger series from Munich and Perth
(P . 0.05).
To corroborate our genetic data, we extended our analysis to
histopathologically confirmed brain samples of subjects with
AD (N ¼ 33) and investigated possible effects of the PLAU
rs2227564 polymorphism on cerebral plaque counts, as one
characteristic feature of AD associated with Ab metabolism.
The brain samples used for this analysis are presented in
Table 4. In a first approach we restricted our analysis to
patients lacking the APOEe4 allele because of its known
effect on plaque pathology. We found significantly higher
plaque counts in temporal cortices of patients carrying the
PLAU rs2227564 T allele (60.3 + 16.9; N ¼ 6) compared
with patients without the rs2227564 T allele (26.3 + 8.8;
N ¼ 9; P ¼ 0.007) (Table 4, Figs 3 and 4). In the complete his-
topathological sample (N ¼ 33), we then determined the influ-
ence of the APOEe4 and PLAU rs2227564 T allele on cerebral
plaque count by linear regression analysis, using age and sex
as covariates. As expected, we found a significant effect for
presence of the APOE e4 allele (P ¼ 0.0069), but also a
weaker independent effect for the PLAU rs2227564 T allele
(P ¼ 0.015). Age and sex, however, did not contribute to the
number of plaques.
PLAU represents a reasonable functional and positional candi-
date gene as it is located under a linkage region of AD on
Chromosome 10q and functionally is involved in the degra-
dation of Ab via plasmin activation.
Table 1. SNP description, allele distribution and association with AD
AD affection allelic P-values (risk allele) OR and (95%CI)
AD complete N ¼ 422EOAD (N ¼ 140)
(onset ?65 years)
LOAD (N ¼ 282)
(onset . 65 years)
rs10762566 75327866 Intergenic
rs2675680 75328870 Intergenic
rs267566375330276 C10orf 55
rs2633298 75336345 Promoter
rs2459449 75339019 Promoter
rs2227551 75339196 Promoter
rs222757975341295 Exon 2
rs2227562 75342967 Intron 5
rs222756475343107 Exon 6
rs2227568 75343885 Exon 8
rs2227582 75344258 Intron 8
rs4065 75346470 Exon 11
rs2461863 75350242 30-UTR
rs2688624 75365730 Intergenic
0.012 (A); 1.34 (1.07–1.67) 0.492
0.032 (C); 1.27 (1.02–1.59)
0.042 (G); 1.26 (1.01–1.57) 0.905
0.019 (G); 1.47 (1.09–1.98) 0.072
0.020 (T); 1.37 (1.05–1.78)
0.033 (C); 1.37 (1.03–1.84)
0.018 (T); 1.31 (1.04–1.65)
0.034 (G); 1.28 (1.02–1.61) 0.813
0.049 (T); 1.26 (1.01–1.58)
0.013 (A); 1.51 (1.1–2.1)
0.01 (C); 1.5 (1.1–2.0)
0.009 (G); 1.5 (1.1–2.0)
0.042 (G); 1.5 (1.01–2.2)
0.001 (T); 1.91 (1.3–2.8) 0.517
0.05 (C); 1.47 (10.99–2.1)
0.007(T) 1.6 (1.2–2.24)
0.008 (G) 1.54 (1.1–2.14)
0.029 (T); 1.42 (1.03–1.95)
0.028 (G); 1.41 (1.04–1.9)
0.011 (G); 1.59 (1.1–2.23)
The locations of the SNPs on Chromosome 10 and PLAU according to the UCSC map (hg17) are shown. The distribution of major allele frequencies in
cases and controls from Munich, the association results (logistic regression including the covariates age at onset/exam and sex) for the complete
sample, patients with early onset AD (EOAD) and late onset AD (LOAD) with P-values and OR with 95%CI for significant SNPs are shown.
n.a., not available.
2450 Human Molecular Genetics, 2006, Vol. 15, No. 16
In an attempt to elucidate the genetic contribution of PLAU
to the development of AD and to further understand the con-
flicting results of several association studies on PLAU, we ana-
lyzed a large case–control sample on a high-density SNP map
using a whole gene approach (24) and analyzed the LD struc-
ture of the genomic region. Using a sliding window approach,
we fine-mapped the associated region. The strongest single
marker association conferring risk to the disease, particular
EOAD was observed for the functional rs2227564 polymorph-
ism, which is located at an LD breakpoint within PLAU Exon
6. Interestingly, the strength of this association increased in
patients with a younger onset of the disease, particularly in
patients with EOAD where we observed a T-allele dose-
dependent increase of the ORs, whereas other, weaker associ-
ation signals appeared in patients with LOAD. Sequencing of
PLAU Exon 6 identified a new functional polymorphism,
which did not contribute to the observed association. As the
rs2227564 polymorphism showed the strongest association
signal and because of its functional nature we replicated this
SNP in three independent case–control series from Bonn,
Brescia and Perth and confirmed our initial results. All
series showed an allelic association and a T-allele dose-
dependent increase of the ORs. In the Australian series the
association appeared, likewise to the Munich sample, to be
stronger in patients with a younger onset of the disease. To
reduce the possibility of a spurious association finding due
to hidden stratification we included a discordant sib-pair
sample from Germany, which also showed a significant over-
representation of the risk allele in the affected group. Because
of the hypothesized functional involvement of PLAU in Ab
catabolism, we finally examined possible effects of the
PLAU rs2227564 T-allele on cerebral plaque counts using
post-mortem brain material of subjects with neuropathological
confirmed AD. Brain samples with the rs2227564 T-allele
showed significantly higher plaque counts compared with non-
carriers, and this effect was independent from the presence of
the APOEe4 allele.
Taken together our results provide considerable evidence of
a genetic and functional involvement of PLAU polymorph-
isms, respectively, of a yet untyped causal variant in LD,
into the pathogenesis of AD.
However, six articles investigating a possible association of
PLAU polymorphisms with AD and related endophenotypes,
such as plasma Ab42 concentrations and cognitive perform-
ance (MMSE) were published so far. In summary, three of
these articles by Myers et al. (19), Papassotiropoulos et al.
(20) and Bagnoli et al. (21) failed to identify a significant
association for the rs2227564 polymorphism with AD.
While Ertekin-Taner et al. (22), reported a similar association
to our study, Finckh et al. (18) observed a risk association to
the opposite allele.
On the first view, these discrepant results do not substan-
tially support a major role of PLAU as a new susceptibility
gene of Alzheimer’s disease, but rather may reflect the
typical situation often observed in genetically complex dis-
eases. Typical reasons for the discrepancies are well known
and include factors such as statistical over-interpretation of
the first study, small sample sizes, genetic and phenotypic het-
erogeneity of the analyzed trait, population differences in
allele frequencies and correlation structures among genetic
markers, drawbacks in study design and differences regarding
ethnic background and population sub-structures (e.g. different
age structures in patients).
Regarding the correlation structures, there is evidence that
the LD structure of PLAU differs substantially among popu-
lation samples. The finding of a LD breakpoint at PLAU
Exon 6 in the German population is supported by a previous
study which examined block boundary shifts in a number of
genes, including PLAU in several European populations
(26). This breakpoint has not been observed in two larger
studies from the US which genotyped five (22), respectively,
seven SNPs (19) within PLAU. Even between European and
German samples, the probability for LD block boundaries
(markers 14 and 22 in figure 2 of Mueller et al.) (26) are
highly variable. Such differences in LD-block structures
including boundary shifts are expected among different popu-
lations and may help to limit association signals to smaller
regions (26). Moreover, the low r2values at the associated
region of PLAU rs2227562 (Intron 5), rs2227564 (Exon 6)
and rs2227568 (Exon 8) and some scattered high r2values
among non-neighboring markers, may indicate a long and
independent mutational history of the individual markers.
Recent mutational events may vary in different populations,
thus leading to different correlation structures between popu-
lations, which complicate replications and comparison of
genetic association studies. However, all polymorphisms in
correlation with our tested coding SNP have to be considered
as equally likely causal candidates. There is a large block
(?300 kb) of scattered correlated SNPs between two indicated
Table 2. Common PLAU haplotypes, frequencies and associations with AD
AD affection P-values
All haplotypes with a frequency .1% within the PLAU gene region (23
SNPs) in the Munich sample were tested for association with AD affec-
tion status using x2statistics.
Human Molecular Genetics, 2006, Vol. 15, No. 162451
encompasses the following 10 genes: FLJ32658, SEC24C,
CAMK2G, C10orf55 and PLAU. On the basis of our results
we assume, however, that our association signal maps to the
PLAU region, which is supported by relatively high, although
variable, probabilities for LD block boundaries between the
neighboring genes CAMK2G and PLAU in most European
samples [around marker 8/9 (rs2675671/rs2633303) and
marker 14 (rs2688611) in figure 2 of Mueller et al.] (26),
which in turn corresponds to higher estimated recombination
rates in the HapMap data. In addition, the 30neighboring
gene VCL is clearly unlinked to PLAU.
Sample size considerations may apply to the studies of
Papassotiropoulos et al. (20) and Finckh et al. (18) who
used small sample sizes, which were less than 100 individ-
uals in some of their analyzed samples (e.g. 43 cases and
55 controls from Basel, Switzerland and 94 patients and
30 controls from Brescia, Italy) (18), or 181 cases and 99
controls from Greece (20). Another weak point of both
studies is the applied pooling strategy, thereby combining
ethnic backgrounds and considerable differences of the geno-
type distributions (upto 10%). In contrast to all other studies,
Finckh et al. (18) identified the major allele to confer risk to
AD. This is of particular interest as we observed a similar
association pattern for both neighboring SNPs (rs2227562
and rs2227568). Although there appears to be large variabil-
ity of the rs2227564 minor allele frequencies among differ-
ent populations with a gradient from Northern Europe
accompanied by an increase of the recombination signal in
the southern parts, the reason for this observation remains
to be elucidated.
An often neglected and underestimated factor that certainly
contributes to the large variability of genetic association
studies concerns sample differences in the age structures.
We also observed only a weak association in our larger
series from Munich and Perth, but subsequent analysis of
the patient substructure revealed that the risk conferred by
rs2227564 appears to be stronger in patients with a younger
onset of AD, whereas no effect was present in elderly patients.
This age-dependency seems reasonable as it may be hypoth-
esized that the genetic contribution, in general and that of a
specific genetic risk factor may vary according to differences
(13%) (26), whichis
Figure 2. Sequencing chromatogram of the PLAU Exon 6 region. Identifi-
cation of SNPs at the PLAU Exon 6 region. One additional coding SNP
was identified by sequencing of genomic DNA in 80 patients with AD and
80 controls (320 alleles).
Table 3. PLAU rs2227564 allele and genotype frequencies in case–control and discordant sib-pair samples
Age at onset/examAllele frequency [%]Genotypes (%), P-values and OR (95%CI)
OR ¼ 1.7 (1.1–2.6)
OR ¼ 3.0 (1.1–8.0)
25768.8+13.1 0.780.22 0.590.380.03
109 74.8+6.4 0.760.24 0.005 0.590.34 0.02
OR ¼ 1.9 (1.1–3.3)
OR ¼ 3.7 (1.1–11.9)
17375.1+8.6 0.85 0.150.730.240.03
120 67.3+8.50.79 0.210.004 0.64 0.300.002
OR¼ 3.2 (1.5–6.9)
OR¼ 8.7 (0.9–82)
AD .80 years
AD ?80 years
9964.6+7.20.92 0.080.860.13 0.01
OR¼ 2.4 (1.4–4.2)
OR¼ 2.9 (1.3–7.1)
33877.1+10.6 0.810.19 0.66 0.290.05
Description of the case–control cohorts from Munich, Bonn, Brescia and Perth and the discordant sib-pair sample with age of onset for patients and age
at exam for controls. Allele frequencies, allelic association results (logistic regression including the covariates age at onset/exam and sex), genotype
frequencies and genotypic association results (logistic regression including the covariates age and sex) with P-values and OR with 95%CI are given for
the four independent case–control series. n.s., not significant.
aStatistical analysis of the discordant sib-pair sample was performed using the S-TDT as described in the methods section.
2452 Human Molecular Genetics, 2006, Vol. 15, No. 16
regarding the onset of the disease. This fact has been repeat-
edly demonstrated for APOE, the strongest genetic risk
factor of AD so far (27). Therefore, studies using relatively
old patient samples, such as Myers et al. (19) (e.g. Mayo
series mean age 82.7 years) may have missed to detect the
possible association with EOAD.
In contrast, Ertekin-Taner et al. (22) demonstrated genetic
results similar to this study in their larger series (MCR)
from the US. They observed not only a significant OR of
1.4 (1.018–1.9) for the functional Exon 6 rs2227564 poly-
morphism, but also significant results for some neighboring
markers at Intron 7, Intron 9 and Exon 11.
Moreover, they present functional data of significantly
elevated plasma Ab42 levels in carriers of the PLAU risk gen-
otypes, which are believed to reflect a quantitative trait of AD.
This result corresponds to our findings of higher cerebral
plaque counts in brain samples with the risk allele. In addition,
they suggest that the elevation of Ab42 levels may be due to a
loss of function of PLAU as aged mice with a knock-out of the
PLAU gene also showed substantially elevated Ab42 levels.
Further support of a functional involvement of PLAU in the
mediation of lethal effects due to APP over-expression
comes from transgenic mouse studies (28). Analysis of age
at death as a quantitative trait in transgenic animals vulnerable
to APP over-expression identified a region on mouse Chromo-
some 14, which counteracts the lethal effects of APP. This
region harbors several sequences with human homologues,
including two genes, PLAU and neuregulin 3 (NRG3) on
Chromosome 10 (28). As preliminary genotyping data from
our group do not support NRG3 as a possible susceptibility
gene of AD (data not shown), it remains to be seen whether
PLAU alone is responsible for the linkage signal.
Table 4. Description of brain samples used for analysis of cerebral plaque counts, description of brain samples and plaque counts of the temporal cerebral cortex
Sex (male/female) Age at death (mean+ SD) Duration (mean+SD) Plaque counts (mean+ SD)
rs2227564 T-allele absent APOEe4 allele absent
rs2227564 T-allele present APOEe4 allele absent
rs2227564 T-allele absent APOEe4 allele present
rs2227564 T-allele present APOEe4 allele present
Numbers, sex distribution, age at death (years, mean +SD), duration of the disease (years, mean +SD) and cortical plaque counts (mean +SD)
according to the presence or absence of the PLAU rs 2227564 T- and the APOEe4 allele are shown. Plaques were counted by examination of
three slides per case and eight consecutive representative fields from severely affected tissue areas.
Figure 3. Plaque counts according to presence or absence of the PLAU s2227564 T and the APOEe4 alleles. Scatter of temporal plaque counts with mean and
95%CI. Grouping with number of samples according to presence or absence of the PLAU rs2227564 T allele and the APOEe4 allele. P-value for group
differences using the two-sided Mann–Whitney U test as indicated.
Human Molecular Genetics, 2006, Vol. 15, No. 162453
Likewise to our study, also Papassotiropoulos et al. (20)
investigated possible functional effects of the rs2227564 poly-
morphism in human brain material but failed to find any
effects. However, the use of brain samples from subjects
without dementia and any neuropathological abnormalities
may be the reason for not identifying any effects on the cer-
ebral plaque load. While Myers et al. (19) excluded variations
in PLAU to contribute to their linkage region at Chromosome
10q (2,10), Ertekin-Taner et al. (22) demonstrated for the
rs2227564 polymorphism, a contribution of 22% to their
plasma Ab42 phenotype (22).
In summary, this study provides compelling evidence of a
genetic and functional involvement of a common PLAU
variant into the pathogenesis of AD. Further functional inves-
tigations are required to elucidate the role of PLAU and PLAU
polymorphisms in the pathogenesis of AD and the processing
of Ab proteins.
MATERIAL AND METHODS
Subjects and material
This study was approved by the review boards of each medical
faculty and refers to a total of 2359 Caucasian subjects com-
promising four independent case–control samples from
Munich, Germany (AD, N ¼ 422; controls ¼ 257); Bonn,
Germany(AD ¼ 109;
(AD ¼ 120; controls ¼ 99); Perth, Australia (AD ¼ 219;
controls ¼ 338) and a discordant sib-pair sample from
Germany (affected ¼ 251; unaffected ¼ 371). All individuals
were recruited from specialists at University Memory Clinics
and each control group was matched for geographical location,
ethnicity, sex and age and consisted of cognitively healthy
individuals. The clinical diagnosis of probable AD was estab-
lished according to National Institute of Neurological and
Communicative disorders and Stroke-Alzheimer’s Disease
and Related Disorder Association (NINCDS-ADRDA) criteria
(29). After informed consent had been obtained, blood
samples of each individual were taken by venous puncture.
Cognitive performance was assessed using standard neuro-
psychological tests, such as the Cambridge Cognitive Examin-
ation (30) or the Consortium to Establish a Registry for
Alzheimer’s disease (31) which includes the Mini Mental
State Examination (MMSE) (32), which was used as a
measure of global cognitive performance in the control
groups. Control subjects with an MMSE score below 28
were excluded from further analyses. All patients and controls
underwent a thorough psychiatric, neurological and neuropsy-
chological evaluation. The diagnostic work-up also included
an informant interview, a chemistry survey and structural
selected patients (11 male, 22 female; mean age at death
78.5 + 7.9 years; disease duration 7.5 + 3.4 years) with clini-
cally and neuropathologically confirmed late-onset AD was
obtained from the Departments of Neuropathology at the
University of Munich and the Technical University Munich.
The neuropathological diagnosis of AD was performed
according to established criteria (33).
controls ¼ 173); Brescia,Italy
Genotyping and sequencing
We genotyped 56 SNPs derived from Public or the Celera
human genome databases covering ?299 kb of the PLAU
gene region on Chromosome 10 including the upstream
located genes calcium/calmodulin-dependent protein kinase
II-gamma (CAMK2G; OMIM # 602123), N-deacetylase/
N-sulfotransferase 2 (NDST2; OMIM # 603268), KIAA0913
and downstream Vinculin (VCL; OMIM # 193065). A total
of 52 SNPs with a median spacing of 6.20kb showed sufficient
genotyping quality (mean call rate above 95%), Hardy–
Weinberg equilibrium and a minor allele frequency of more
than 0.05 in the controls (Fig. 1). At the PLAU region, we gen-
otyped 24 markers with a mean intermarker distance of 2 kb.
The markers and locations are described in detail in Table 1.
APOE genotypes were determined by the restriction enzyme
approach (34). Genotyping was performed at the Department
of Psychiatry, TU Munich and the Genome Analysis Center,
GSF, Munich by a primer extension of multiplex polymerase
Figure 4. Immunohistochemistry of plaques according to the PLAU rs2227564
alleles. Comparison of Ab load of the temporal cortex in two cases lacking the
APOEe4 allele with (A) and without (B) the PLAU rs2227564 T allele. Immu-
nohistochemistry with monoclonal Ab antibody (DAKO; 6F/3D; 1:100),
peroxidase/DAB. Scale bar: 500 mm.
2454Human Molecular Genetics, 2006, Vol. 15, No. 16
chain reaction products and detection of the allele-specific
extension products by matrix-associated laser desorption/
ionization time of flight (MALDI-TOF) mass spectrometry
(Sequenom, San Diego, USA).
We sequenced genomic DNA (PLAU Exon 6 including the
Exon/Intron boundaries at both sides) from 80 patients with
AD and 80 cognitively healthy controls using Applied Biosys-
tems BigDye Chemistry according to the recommendations of
DNA from paraffin-embedded brain tissues was extracted as
described (35), whereas standard methodologies were used for
frozen brain samples. APOE and PLAU genotypes of brain
the MALDI-TOF and the restriction enzyme approach (34,36).
Analysis of cerebral plaque counts and
The quantitative analysis of plaques and Ab immunohisto-
chemistry was performed using sections from the medial tem-
poral gyrus according to the methodology described before
(37,38). In particular, the number of plaques were counted
by an experienced neuropathologist who was blinded for the
genotyping results by examination of three slides per case
and eight consecutive representative fields from severely
affected tissue areas.
The pairwise linkage disequilibrium measures, D0and r2, were
calculated using the software package Haploview (39).
Association between AD affection state and the alleles/
genotypes of SNP markers was tested by logistic regression
analysis including age of onset/examination and sex as covari-
ates. Corrections for multiple comparisons were considered
using a global permutation test (1000 permutations). Multipli-
cative interaction terms were tested between SNP effects and
APOEe4 status (present or absent), age and gender by appro-
priate logistic regression models. Haplotype assignments for
each individual were estimated by the EM algorithm and sub-
sequently used for association tests, both implemented in
Haploview (39). A sliding window approach was performed
to test for all two-, and three-locus haplotype associations.
The discordant sib-pair sample was analyzed using the sib
transmission/disequilibrium test (S-TDT) (25).
Differences in plaque counts were analyzed according to the
presence or absence of the PLAU risk alleles by the two-sided
Mann–Whitney U test. An overall analysis of genetic effects
on cerebral plaque pathology was estimated by linear
regression analysis using the APOEe4 allele, PLAU risk
allele, duration of the disease and gender as covariates.
Supplementary Material is available at HMG Online.
We wish to thank all individuals who participated in this
study for their contribution and would also like to thank the
dementia outpatient unit employees and Tamara Eisele for
their help in collecting and processing of samples. This
work was funded by the German National Genome Network
(NGFN) and the German Ministry for Education and
Research; Grant Numbers 01GS0465 to M.R., 01GS0166
to M.R., A.K., U.M., 01GS0116 to H.B. and FE 77443
to T.I. R.N.M. and K.T. were supported by The McCusker
Foundation for Alzheimer’s Disease Research.
Conflict of Interest statement. None of the authors has had
involvements that might raise the question of bias in the
work reported or in the conclusions, implications or opinions
1. Hardy, J. and Selkoe, D.J. (2002) The amyloid hypothesis of Alzheimer’s
disease: progress and problems on the road to therapeutics. Science, 297,
2. Myers, A., De-Vrieze, F.W., Holmans, P., Hamshere, M., Crook, R.,
Compton, D., Marshall, H., Meyer, D., Shears, S., Booth, J. et al. (2002)
Full genome screen for Alzheimer disease: stage II analysis. Am. J. Med.
Genet., 114, 235–244.
3. Scott, W.K., Hauser, E.R., Schmechel, D.E., Welsh-Bohmer, K.A.,
Small, G.W., Roses, A.D., Saunders, A.M., Gilbert, J.R., Vance, J.M.,
Haines, J.L. et al. (2003) Ordered-subsets linkage analysis detects novel
Alzheimer disease loci on Chromosomes 2q34 and 15q22. Am. J. Hum.
Genet., 73, 1041–1051.
4. Li, Y.J., Scott, W.K., Hedges, D.J., Zhang, F., Gaskell, P.C., Nance, M.A.,
Watts, R.L., Hubble, J.P., Koller, W.C., Pahwa, R. et al. (2002) Age at
onset in two common neurodegenerative diseases is genetically
controlled. Am. J. Hum. Genet., 70, 985–993.
5. Blacker, D., Bertram, L., Saunders, A., Moscarillo, T., Albert, M.,
Wiener, H., Perry, R., Collins, J., Harrell, L., Go, R. et al. (2003) Results
of a high resolution genome screen of 437 Alzheimer’s disease families.
Hum. Mol. Genet., 12, 23–32.
6. Holmans, P., Hamshere, M., Hollingworth, P., Rice, F., Tunstall, N.,
Jones, S., Moore, P., Wavrant DeVrieze, F., Myers, A., Crook, R. et al.
(2005) Genome screen for loci influencing age at onset and rate of decline
in late onset Alzheimer’s disease. Am. J. Med. Genet. B Neuropsychiatr.
Genet., 135, 24–32.
7. Wijsman, E.M., Daw, E.W., Yu, C.E., Payami, H., Steinbart, E.J.,
Nochlin, D., Conlon, E.M., Bird, T.D. and Schellenberg, G.D. (2004)
Evidence for a novel late-onset Alzheimer disease locus on Chromosome
19p13.2. Am. J. Hum. Genet., 75, 398–409.
8. Corder, E.H., Saunders, A.M., Strittmatter, W.J., Schmechel, D.E.,
Gaskell, P.C., Small, G.W., Roses, A.D., Haines, J.L. and
Pericak-Vance, M.A. (1993) Gene dose of apolipoprotein E type 4 allele
and the risk of Alzheimer’s disease in late onset families. Science, 261,
Goldfarb, L., Goldgaber, D., Manwaring, M.G., Szymanski, M.H.,
McCown, N. et al. (1993) Apolipoprotin E e4 allele distributions in
late-onset Alzheimer’s disease and in other amyloid-forming diseases.
10. Myers, A., Holmans, P., Marshall, H., Kwon, J., Meyer, D., Ramic, D.,
Shears, S., Booth, J., DeVrieze, F., Crook, R. et al. (2000) Susceptibility
locus for Alzheimer’s disease on Chromosome 10. Science, 290,
11. Bertram, L., Blacker, D., Mullin, K., Keeney, D., Jones, J., Basu, S.,
Yhu, S., McInnis, M., Go, R., Vekrellis, K. et al. (2000) Evidence for
genetic linkage of Alzheimer’s disease to Chromosome 10q. Science, 290,
Adamson, J., Ronald, J., Blangero, J., Hutton, M. and Younkin, S. (2000)
Linkage of plasma Ab42 to a quantitative locus on Chromosome 10 in late
onset Alzheimer’s disease pedigrees. Science, 290, 2303–2304.
13. Davis, J., Wagner, M.R., Zhang, W., Xu, F. and Van Nostrand, W.E.
(2003) Amyloid beta-protein stimulates the expression of urokinase-type
Human Molecular Genetics, 2006, Vol. 15, No. 16 2455
plasminogen activator (uPA) and its receptor (uPAR) in human
cerebrovascular smooth muscle cells. J. Biol. Chem., 278, 19054–19061.
14. Walker, D.G., Lue, L.F. and Beach, T.G. (2002) Increased expression of
the urokinase plasminogen-activator receptor in amyloid beta
peptide-treated human brain microglia and in AD brains. Brain. Res.,
15. Tucker, H., Kihiko, M., Caldwell, J., Wright, S., Kawarabayashi, T.,
Price, D., Walker, D., Scheff, S., McGillis, J., Rydel, R. et al. (2000) The
plasmin system is induced by and degrades Amyloid-b aggregates.
J. Neurosci., 20, 3937–3946.
16. Tucker, H., Kihiko-Ehmann, M. and Estus, S. (2002) Urokinase-type
plasminogen activator inhibits amyloid-beta neurotoxicity and
fibrillogenesis via plasminogen. J. Neurosci. Res., 70, 249–255.
17. Ledesma, M.D., Da Silva, J.S., Crassaerts, K., Delacourte, A., De
Strooper, B. and Dotti, C.G. (2000) Brain plasmin enhances APP
alpha-cleavage and Abeta degradation and is reduced in Alzheimer’s
disease brains. EMBO Rep., 1, 530–535.
18. Finckh, U., vanHadeln, K., Mu ¨ller-Thomsen, T., Alberici, A., Binetti, G.,
Hock, C., Nitsch, R., Stoppe, G., Reiss, J. and Gal, A. (2003) Association
of late-onset Alzheimer disease with a genotype of PLAU, the gene
encoding urokinase-type plasminogen activator on Chromosome 10q22.2.
Neurogenetics, 4, 213–217.
19. Myers, A.J., Marshall, H., Holmans, P., Compton, D., Crook, R.J.,
Mander, A.P., Nowotny, P., Smemo, S., Dunstan, M., Jehu, L. et al.
(2004) Variation in the urokinase-plasminogen activator gene does not
explain the Chromosome 10 linkage signal for late onset AD. Am. J. Med.
Genet. B Neuropsychiatr. Genet., 124, 29–37.
Huynh, K.D., Tracy, J., Staehelin, H.B., Monsch, A.U., Nitsch, R.M. et al.
(2005) No association of a non-synonymous PLAU polymorphism with
Alzheimer’s disease and disease-related traits. Am. J. Med. Genet. B
Neuropsychiatr. Genet., 132, 21–23.
21. Bagnoli, S., Tedde, A., Cellini, E., Rotondi, M., Nacmias, B. and Sorbi, S.
(2005) The urokinase-plasminogen activator (PLAU) gene is not
associated with late onset Alzheimer’s disease. Neurogenetics, 6, 53–54.
22. Ertekin-Taner, N., Ronald,J., Feuk,L., Prince,J., Tucker, M., Younkin,L.,
Hella, M., Jain, S., Hackett, A., Scanlin, L. et al. (2005) Elevated amyloid
with single nucleotide polymorphisms in the urokinase-type plasminogen
activator gene. Hum. Mol. Genet., 14, 447–460.
23. Blomqvist, M.E., Reynolds, C., Katzov, H., Feuk, L., Andreasen, N.,
Bogdanovic, N., Blennow, K., Brookes, A.J. and Prince, J.A. (2006)
Towards compendia of negative genetic association studies: an example
for Alzheimer disease. Hum. Genet., 119, 29–37.
24. Neale, B.M. and Sham, P.C. (2004) The future of association studies:
gene-based analysis and replication. Am. J. Hum. Genet., 75, 353–362.
25. Spielman, R. and Ewens, W. (1998) A sibship test for Linkage in the
presence of association: the sib transmission/disequilibrium test.
Am. J. Hum. Genet., 62, 450–458.
26. Mueller, J.C., Lohmussaar, E., Magi, R., Remm, M., Bettecken, T.,
Lichtner, P., Biskup, S., Illig, T., Pfeufer, A., Luedemann, J. et al. (2005)
Linkage disequilibrium patterns and tagSNP transferability among
European populations. Am. J. Hum. Genet., 76, 387–398.
27. Bickeboller, H., Campion, D., Brice, A., Amouvel, P., Hannequin, D.,
Didierjean, O., Penet, C., Martin, C., Perez-Tur, J., Michon, A. et al.
(1997) Apolipoprotein E and Alzheimer’s disease: genotype specific risks
by age and sex. Am. J. Hum. Genet., 60, 439–446.
Westaway, D., Younkin, L., Younkin, S.G., Ashe, K.H. et al. (2004)
Identification of loci determining susceptibility to the lethal effects of
29. McKhann, G., Folstein, M., Katzman, R., Price, D. and Stadlan, E.M.
(1984) Clinical diagnosis of Alzheimer’s disease: Report of the
NINCDS-ADRDA work group under the auspices of Department of
Health and Human Services Task Force on Alzheimer’s Disease.
Neurology, 34, 939–944.
30. Roth, M., Tym, E., Mountjoy, C.Q., Huppert, F.A., Hendrie, H., Verma, S.
and Goddard, R. (1986) CAMDEX. A standardized instrument for the
diagnosis of mental disorder in the elderly with special reference to the
early detection of dementia. Br. J. Psychiat., 149, 698–709.
31. Welsh, K.A., Butters, N., Mohs, R.C., Beekly, D., Edland, S.,
Fillenbaum, G. and Heyman, A. (1994) The Consortium to Establish a
Registry for Alzheimer’s Disease (CERAD). Part V. A normative study of
the neuropsychological battery. Neurology, 44, 609–614.
32. Folstein, M.F., Folstein, S.E. and McHugh, P.R. (1975) ‘Mini Mental
State0. A practical method for grading the cognitive state of patients for
the clinician. J. Psychiat. Res., 12, 189–198.
33. Mirra, S.S., Heyman, A., McKeel, D., Sumi, S.M., Crain, B.J.,
Brownlee, L.M., Vogel, F.S., Hughes, J.P., van-Belle, G. and Berg, L.
(1991) The Consortium to Establish a Registry for Alzheimer’s Disease
(CERAD). Part II. Standardization of the neuropathologic assessment of
Alzheimer’s disease. Neurology, 41, 479–486.
34. Zivelin, A., Rosenberg, N., Peretz, H., Amit, Y., Kornbrot, N. and
Seligsohn, U. (1997) Improved method for genotyping apolipoprotein E
polymorphisms by a PCR-based assay simultaneously utilizing two
distinct restriction enzymes. Clin. Chem., 43, 1657–1659.
35. Ko ¨sel, S. and Graeber, M.B. (1994) Use of neuropathological tissue for
molecular genetic studies: parameters affecting DNA extraction and
polymerase chain reaction. Acta Neuropathol., 88, 19–25.
36. Tu ¨rkmen, B., Schmitt, M., Schmalfeldt, B., Trommler, P., Hell, W.,
Creutzburg, S., Graeff, H. and Magdolen, V. (1997) Mutational analysis of
the genes encoding urokinase-type plasminogen activator (uPA) and its
inhibitor PAI-1 in advanced ovarian cancer. Electrophoresis, 18,
37. Schmechel, D.E.A., Saunders, A.M., Strittmatter, W.J., Crain, B.J.,
Hulette, C.M., Joo, S.H., Pericak-Vance, M.A., Goldgaber, D. and
Roses, A.D. (1993) Increased amyloid b-peptide deposition in cerebral
cortex as a consequence of apolipoprotein E genotype in late-onset
Alzheimer’s disease. Proc. Natl Acad. Sci. USA, 90, 9649–9653.
38. Egensperger, R., Ko ¨sel, S., Eitzen, U.v. and Graeber, M. (1998)
Microglial activation in Alzheimer’s disease: Association with APOE
genotype. Brain Pathol., 8, 439–447.
39. Barrett, J.C., Fry, B., Maller, J. and Daly, M.J. (2005) Haploview: analysis
and visualization of LD and haplotype maps. Bioinformatics, 21,
The German Sib-Pair Study Group: Many data and biomater-
ials were collected by two recruitment sides of the German
Sib-Pair Study Group at the TU-Munich (M. Kro ¨mmer,
B. Cramer, A. Klimbacher), the GSF, Munich (P. Belcredi)
and several centers that participated at this study. The princi-
pal investigators and co-investigators were: Bezirkskranken-
haus Gabersee; Professor Dr G. Laux and Dr Eberl;
Department of Psychiatry, University of Regensburg, Pro-
fessor Dr H.E. Klein and Dr Bernd Ibach; Neurologische
Klinik Bad Aibling, Professor Dr Eberhard Koenig and Dr
Barbara Romero; Bezirkskrankenhaus Bayreuth, Professor
Dr Manfred Wolfersdorf and Dr Michael Schu ¨ler; Department
of Psychiatry, University of Freiburg, Professor Dr Mathias
Berger, PD Dr Schmidtke; E. Jost MSc; Bezirkskrankenhaus
Augsburg Professor Dr M. Schmauss and Ch. Steber;
Krankenhaus Mu ¨nchen-Neuperlach, Abt. fu ¨r Akutgeriatrie,
Professor Dr R. Heinrich and Dr Britta Wiegele; Bezirkskran-
kenhaus Werneck PD Dr H.-P. Volz and Dr M. Ja ¨hnel;
Bezirkskrankenhaus Taufkirchen, Professor Dr M. Dose, Dr
Marquard and Dr Bremer; Psychiatrische Klinik Agatharied,
Dr N. Braunischand Dr
Mainkofen, Dr L. Blaha and Dr S. Herpich; Bezirkskranken-
haus Haar, Dr H.W. Dietl; Bezirkskrankenhaus Gu ¨nzburg,
Professor Dr R. Schu ¨ttler and PD Dr R. Hess; Bezirkskranken-
haus Landshut, Professor Dr M. Philipp and Dr A. Wermuth.
2456Human Molecular Genetics, 2006, Vol. 15, No. 16