A Scan of Chromosome 10 Identifies a Novel Locus Showing Strong Association with Late-Onset Alzheimer Disease

Article (PDF Available)inThe American Journal of Human Genetics 78(1):78-88 · January 2006with51 Reads
DOI: 10.1086/498851 · Source: PubMed
Abstract
Strong evidence of linkage to late-onset Alzheimer disease (LOAD) has been observed on chromosome 10, which implicates a wide region and at least one disease-susceptibility locus. Although significant associations with several biological candidate genes on chromosome 10 have been reported, these findings have not been consistently replicated, and they remain controversial. We performed a chromosome 10-specific association study with 1,412 gene-based single-nucleotide polymorphisms (SNPs), to identify susceptibility genes for developing LOAD. The scan included SNPs in 677 of 1,270 known or predicted genes; each gene contained one or more markers, about half (48%) of which represented putative functional mutations. In general, the initial testing was performed in a white case-control sample from the St. Louis area, with 419 LOAD cases and 377 age-matched controls. Markers that showed significant association in the exploratory analysis were followed up in several other white case-control sample sets to confirm the initial association. Of the 1,397 markers tested in the exploratory sample, 69 reached significance (P < .05). Five of these markers replicated at P < .05 in the validation sample sets. One marker, rs498055, located in a gene homologous to RPS3A (LOC439999), was significantly associated with Alzheimer disease in four of six case-control series, with an allelic P value of .0001 for a meta-analysis of all six samples. One of the case-control samples with significant association to rs498055 was derived from the linkage sample (P = .0165). These results indicate that variants in the RPS3A homologue are associated with LOAD and implicate this gene, adjacent genes, or other functional variants (e.g., noncoding RNAs) in the pathogenesis of this disorder.
78 The American Journal of Human Genetics Volume 78 January 2006 www.ajhg.org
A Scan of Chromosome 10 Identifies a Novel Locus Showing Strong
Association with Late-Onset Alzheimer Disease
Andrew Grupe,
1
Yonghong Li,
1
Charles Rowland,
1
Petra Nowotny,
2
Anthony L. Hinrichs,
2
Scott Smemo,
2
John S. K. Kauwe,
2
Taylor J. Maxwell,
2
Sara Cherny,
2
Lisa Doil,
1
Kristina Tacey,
1
Ryan van Luchene,
1
Amanda Myers,
3
Fabienne Wavrant-De Vrie`ze,
3
Mona Kaleem,
3
Paul Hollingworth,
4
Luke Jehu,
4
Catherine Foy,
5
Nicola Archer,
5
Gillian Hamilton,
5
Peter Holmans,
4
Chris M. Morris,
6
Joseph Catanese,
1
John Sninsky,
1
Thomas J. White,
1
John Powell,
5
John Hardy,
3
Michael O’Donovan,
4
Simon Lovestone,
5
Lesley Jones,
4
John C. Morris,
2
Leon Thal,
7
Michael Owen,
4
Julie Williams,
4
and Alison Goate
2
1
Celera Diagnostics, Alameda, CA;
2
Departments of Psychiatry, Neurology, Biology, and Genetics, Washington University, St. Louis;
3
National
Institute on Aging (NIA), Bethesda;
4
Biostatistics and Bioinformatics Unit and Department of Psychological Medicine, Wales College of
Medicine, Cardiff University, Cardiff;
5
Department of Neuroscience, Institute of Psychiatry, King’s College London, London;
6
Institute for
Ageing and Health, Newcastle General Hospital, Newcastle upon Tyne, United Kingdom; and
7
Department of Neurosciences, University of
California–San Diego, La Jolla
Strong evidence of linkage to late-onset Alzheimer disease (LOAD) has been observed on chromosome 10, which
implicates a wide region and at least one disease-susceptibility locus. Although significant associations with several
biological candidate genes on chromosome 10 have been reported, these findings have not been consistently rep-
licated, and they remain controversial. We performed a chromosome 10–specific association study with 1,412 gene-
based single-nucleotide polymorphisms (SNPs), to identify susceptibility genes for developing LOAD. The scan
included SNPs in 677 of 1,270 known or predicted genes; each gene contained one or more markers, about half
(48%) of which represented putative functional mutations. In general, the initial testing was performed in a white
case-control sample from the St. Louis area, with 419 LOAD cases and 377 age-matched controls. Markers that
showed significant association in the exploratory analysis were followed up in several other white case-control
sample sets to confirm the initial association. Of the 1,397 markers tested in the exploratory sample, 69 reached
significance ( ). Five of these markers replicated at in the validation sample sets. One marker, rs498055,P
! .05 P ! .05
located in a gene homologous to RPS3A (LOC439999), was significantly associated with Alzheimer disease in four
of six case-control series, with an allelic P value of .0001 for a meta-analysis of all six samples. One of the case-
control samples with significant association to rs498055 was derived from the linkage sample ( ). TheseP p .0165
results indicate that variants in the RPS3A homologue are associated with LOAD and implicate this gene, adjacent
genes, or other functional variants (e.g., noncoding RNAs) in the pathogenesis of this disorder.
Received August 9, 2005; accepted for publication October 11, 2005; electronically published November 15, 2005.
Address for correspondence and reprints: Dr. Alison Goate, Department of Psychiatry, B8134, Washington University School of Medicine
660 S. Euclid Avenue, St. Louis, MO 63110. E-mail: goate@icarus.wustl.edu
Am. J. Hum. Genet. 2006;78:78–88. 2005 by The American Society of Human Genetics. All rights reserved. 0002-9297/2006/7801-0009$15.00
Alzheimer disease (AD [MIM 104300]) is the most sig-
nificant cause of dementia in developed countries and
is clinically characterized by memory loss of subtle on-
set followed by a slowly progressive dementia that has
a course of several years. The risk of AD has a genetic
component, as evidenced by an increased risk of AD
among first-degree relatives of affected individuals. So far,
three genes have been identified that lead to the rare au-
tosomal dominant early-onset form of AD. Mutations in
the three genes—b-amyloid precursor protein (APP [MIM
104760]) (Goate et al. 1991), presenilin 1 (PSEN1 [MIM
104311]) (Sherrington et al. 1995), and presenilin 2
(PSEN2 [MIM 600759]) (Levy-Lahad et al. 1995)—lead
to an increase in the production of long amyloid b peptide
(Ab42), the main component in amyloid plaques. The
great majority of AD cases are of late onset (age at onset
165 years) and show complex, non-Mendelian patterns
of inheritance. Late-onset AD (LOAD [MIM 606626])
probably results from the combined effects of variation
in a number of genes as well as from environmental fac-
tors. Early genetic studies of LOAD demonstrated that
the 4 variant of APOE (MIM 107741) is associated with
increased risk of LOAD and with lower age at disease
onset in a dose-dependent manner (Corder et al. 1993).
Genomewide linkage screens in patients with LOAD
have identified several other chromosomal regions (re-
viewed by Pastor and Goate [2004]), implying that ge-
netic risk factors other than APOE must exist. Putative
LOAD-susceptibility loci on chromosomes 9, 10, and 12
have been reported in two or more sample sets by dif-
www.ajhg.org Grupe et al.: Association in AD 79
Figure 1 Allelic P values of 1,397 exploratory markers from the
exploratory sample (middle), with a bar graph showing the distribution
of annotated genes across chromosome 10 (bottom). Marker rs498055
is noted with an arrow, and a P value of .05 is marked with a line.
The previously identified linkage peak regions are noted with solid
lines and references (top). Studies with multipoint LOD scores
12in
white samples were included. Results of single-marker studies were
not included.
ferent groups (Pericak-Vance et al. 1997, 2000; Rogaeva
et al. 1998; Kehoe et al. 1999; Myers et al. 2000, 2002;
Blacker et al. 2003). Perhaps the most prominent among
them is the linkage to chromosome 10, observed in a
number of nonoverlapping samples from studies em-
ploying distinct approaches, including linkage analysis
based on a genomewide screen, a candidate gene–based
limited genome screen, and a genome screen that used
plasma Ab levels as a quantitative phenotype (Kehoe et
al. 1999; Bertram et al. 2000; Ertekin-Taner et al. 2000;
Myers et al. 2000; Blacker et al. 2003; Farrer et al.
2003). Several candidate genes that are under or near
the chromosome 10 linkage peaks have been tested for
association with LOAD, but none has been consistently
replicated (Alzheimer Disease Forum).
To identify the genes and mutations for LOAD, we
undertook a screen of putative functional SNPs in 677
genes under the linkage peak, using a powerful set of
unrelated cases and controls. A similar approach was
used to identify the glyceraldehyde-3-phosphate dehy-
drogenase gene (GAPD [MIM 138400]), located on the
short arm of chromosome 12, as a putative LOAD risk
gene (Li et al. 2004). Here, we report the findings from
this scan of 1,412 SNPs on chromosome 10.
Material and Methods
Sample-Set Characteristics
Three white clinical case-control series were used in this study:
(1) the WU series (422 cases; 382 controls), collected through
the Washington University Alzheimer’s Disease Research Cen-
ter (ADRC) patient registry; (2) the UK series (368 cases; 404
controls), collected as part of the Medical Research Council
(MRC) Late-Onset AD Genetic Resource, including those from
the Cardiff University Wales School of Medicine and from King’s
College London; and (3) the UCSD series (217 cases; 409 con-
trols), collected through the ADRC of the University of Cal-
ifornia–San Diego. In total, 1,007 AD cases and 1,195 controls
were analyzed. Cases in these series had received a clinical
diagnosis of dementia of the Alzheimer type (DAT), with use
of criteria equivalent to NINCDS-ADRDA (National Insti-
tute of Neurological and Communicative Diseases and Stroke/
Alzheimer’s Disease and Related Disorders Association) (Mc-
Khann et al. 1984) but modified slightly to include AD as a
diagnosis for individuals aged
190 years (Berg et al. 1998).
The minimum age at onset of DAT was 60 years. Controls
were nondemented individuals aged
160 years at assessment
who were screened for dementia through use of neuropsycho-
logical tests and clinical interviews. Controls were matched
with cases for age and sex. These samples all show an expected
age and APOE 4–genotype distribution and do not appear
to have evidence of population stratification (Li et al. 2004).
More-detailed information about these samples can be found
elsewhere (Li et al. 2005).
A fourth case-control series was generated by selecting one
case per family from our genetic linkage sample (Myers et al.
2002) and matching each of them to an equal number of white,
nondemented controls collected in St. Louis (these controls are
independent of the controls used in the exploratory sample
above). There were 429 cases and 321 controls in this series
(mean age at onset for the case series is 73.6 years; mean age
at assessment for controls is 75.0 years). The linkage pedigrees
from the National Institute of Mental Health (NIMH) series
and the NIA series (292 pedigrees; 624 affected individuals)
(Myers et al. 2002) were also genotyped for the single SNP
significant in all case-control series.
Two small series that consisted of neuropathologically con-
firmed white cases and controls were derived from the U.S.
ADRCs (contributing centers are listed in the Acknowledg-
ments) and from Newcastle upon Tyne, United Kingdom. Of
the samples in the U.S. series, 40% were assessed as being at
either Braak and Braak stage 5 or 6 (cases) or Braak stage 2
or less (controls). The remaining samples (cases) met neuro-
pathological criteria for AD. Both the controls and cases were
selected to be largely free of such complicating pathologies as
Lewy bodies and vascular events. The combined series included
360 cases (age range 65–97 years; 220 women) and 252 con-
trols (age range 65–100 years; 123 women).
SNP Selection and Genotyping
Genotyping of all samples was performed with written in-
formed consent/assent from the participating individuals and
their caregivers and approval from the participating institu-
tions. Polymorphisms used for genotyping were identified from
either the Celera human genome database that includes pub-
licly available SNP data or the Applera Genome Sequencing
Initiative database. For this study, we chose gene-based SNPs,
with a preference for putative functional mutations, as pre-
dicted in the Celera or public SNP databases, with the aim to
screen as many predicted genes with at least one variant as
possible (table A1 and fig. A1 [online only]). Thus, these SNPs
consist of 367 missense/nonsense mutations, 1 donor splice-
80 The American Journal of Human Genetics Volume 78 January 2006 www.ajhg.org
Figure 2 Allelic P values of markers around the RPS3A homologue region (LOC43999) in both exploratory and validation samples,
along with a gene map of the region and Celera assembly coordinates (in Mbp). Blue diamonds indicate two-sided explatory sample P values;
the other symbols indicate one-sided replication sample P values for WU (red squares), UCSD (gray triangles), and UK (green circle).
site variant, 172 putative transcription factor binding site mu-
tations, 9 exon-skipping site variants, 109 variants in the UTR,
and 739 variants of other types (intronic, silent, and unknown
types [SNPs of unknown and silent types were annotated as
functional variants in previous genome assemblies]). They cover
a total of 677 of 1,270 annotated genes on chromosome 10.
All genomic positions for all SNPs and genes are from the
Celera Genome Assembly R27. All SNPs had a minor-allele
frequency (MAF) of
12% in either cases or controls. The MAF
was 2%–5% for 80 exploratory markers and 5%–10% for
165 markers. The remaining markers had MAFs of 10%–
50%, with approximately equal numbers of SNPs in each 10%
interval.
Genotyping of SNPs was undertaken by allele-specific real-
time PCR for individual samples, by use of primers designed
and validated in-house (Germer et al. 2000). Cases and con-
trols were always run on the same plate in a blind fashion.
Assay quality was scored by an individual who had no access
to the sample phenotypes, before the genotyping results were
subjected to statistical analysis. Overall, the accuracy of our
genotyping was
199%, as determined by internal comparisons
of differentially designed assays for the same marker and by
comparisons of the same marker across different groups.
Genotyping was performed in stages—markers were first
genotyped in one sample set, the exploratory set. Generally,
the WU sample set was used as the exploratory sample set.
However, when markers were tested for replication in another
sample set, we also genotyped that sample set with novel assays
that had passed our assay-validation step. Overall, we used
the UK sample (105 assays) and the UCSD sample (1 assay)
as exploratory sets for
!8% of all tested assays. Significant
exploratory markers ( ) were then genotyped in two ad-P
! .05
ditional clinical case-control series. After replication in at least
one of these other sample sets, additional fine-mapping mark-
ers were genotyped near the replicated SNPs. When additional
assays for markers near significant exploratory markers were
immediately available, they were genotyped in the exploratory
sample in parallel with attempting to replicate in the validation
samples. Significant markers were followed up as described
above. Five SNPs that showed some level of replication in one
or both of the additional case-control series were genotyped
in a case-control series derived from the families originally used
for our genomewide linkage scan. One of these SNPs (rs498055)
was also genotyped in the case-control series that was com-
posed of neuropathologically confirmed AD cases and controls.
Statistical Analysis
To help exclude assays with possible genotyping errors from
the analysis, Hardy-Weinberg equilibrium tests were first per-
formed in both the case and the control samples. Assays with
significant deviation from Hardy-Weinberg equilibrium in con-
trols were then examined for genotyping quality ( ; 63P
! .05
markers in the exploratory stage). As a result, two assays were
dropped from our analysis. One remaining assay with an MAF
!10% was significant in the exploratory set but did not validate
in the other sample sets and was in Hardy-Weinberg equilibrium.
Pearson’s x
2
test was used to calculate P values for the as-
sociation of an allele with disease status within a single study.
This test of association was performed on the basis of the fre-
quency counts of a contingency table of allele and disease2 # 2
status. Two-sided P values are presented for the exploratory
study. In the validation stage, one-sided P values were calcu-
lated if the odds ratios (ORs) were in the direction observed
in the exploratory stage. P values were not adjusted for mul-
tiple comparisons unless otherwise stated. ORs and the 95%
CIs for an allelic effect were also estimated. ORs and P values
for meta-analyses that combine results of multiple sample sets
were calculated using the Cochran-Mantel-Haenszel test, and
were controlled for the sample set (Agresti 1990). Evidence of
Table 1
Allelic Tests of Replicated Markers and LOAD
M
ARKER
,G
ENE
,P
OSITION
(bp)
AND
S
AMPLE
C
ASES
C
ONTROLS
MAF
(%) P OR (CI
a
)
P
OWER
TO
R
EPLICATE
No. with Genotype
MAF
(%)
No. with Genotype
11 12 22 Total 11 12 22 Total
rs1057971, PCGF5, 86733401:
WU
b
2 54 363 419 6.9 0 33 344 377 4.4 .029 1.62 (1.05–.52)
UCSD
c
2 33 213 248 7.5 2 37 360 399 5.1 .044 1.49 (1.01–.19) .61
UK
c
1 39 307 347 5.9 3 32 345 380 5.0 .22 1.19 (.82–.75) .66
UCSD and UK
c
3 72 520 595 6.6 5 69 705 779 5.1 .042 1.33 (1.01–.74) .88
All 5 126 883 1,014 6.7 5 102 1,049 1,156 4.8 .0068 1.43 (1.10–.85)
rs498055, LOC439999, 91096111:
UK
b
80 175 92 347 48.3 67 194 124 385 42.6 .029 1.26 (1.02–.55)
UCSD
c
64 107 48 219 53.7 85 156 102 343 47.5 .022 1.28 (1.04–.56) .59
WU
c
125 175 89 389 54.6 65 200 86 351 47.0 .0017 1.36 (1.14–.61) .71
UCSD and WU
c
189 282 137 608 54.3 150 356 188 694 47.3 .00021 1.32 (1.16–.51) .89
All 269 457 229 955 52.1 217 550 312 1,079 45.6 .00004 1.3 (1.15–.47)
rs4417206, ALDH18A1, 91137678:
UK
b
36 153 158 347 32.4 61 169 154 384 37.9 .029 .79 (.63–.98)
UCSD
c
17 102 101 220 30.9 50 157 142 349 36.8 .021 .77 (.62–.95) .59
WU
c
45 155 190 390 31.4 38 165 148 351 34.3 .12 .88 (.73–.05) .71
UCSD and WU
c
62 257 291 610 31.2 88 322 290 700 35.6 .013 .83 (.72–.95) .90
All 98 410 449 957 31.7 149 491 444 1,084 36.4 .0019 .81 (.71–.93)
rs600879, SORCS1, 102662200:
WU
b
6 87 325 418 11.8 4 54 319 377 8.2 .017 1.5 (1.07–.09)
UCSD
c
3 42 196 241 10.0 6 53 335 394 8.2 .15 1.23 (.89–.71) .67
UK
c
6 66 277 349 11.2 3 63 319 385 9.0 .079 1.28 (.96–.70) .74
UCSD and UK
c
9 108 473 590 10.7 9 116 654 779 8.6 .040 1.26 (1.01–.56) .92
All 15 195 798 1,008 11.2 13 170 973 1,156 8.5 .0043 1.34 (1.10–.65)
rs1903908, hCG2039140, 102940843:
WU
b
14 97 308 419 14.9 5 77 294 376 11.6 .050 1.34 (1.00–.80)
UCSD
c
6 54 188 248 13.3 2 81 314 397 10.7 .079 1.28 (.96–.71) .53
UK
c
11 78 247 336 14.9 10 76 296 382 12.6 .10 1.22 (.94–.57) .57
UCSD and UK
c
17 132 435 584 14.2 12 157 610 779 11.6 .029 1.24 (1.03–.50) .80
All 31 229 743 1,003 14.5 17 234 904 1,155 11.6 .0070 1.28 (1.07–.53)
a
95% CI for exploratory and total samples; 90% CI for validation samples.
b
Exploratory sample set.
c
Validation sample set: one-sided P value.
82 The American Journal of Human Genetics Volume 78 January 2006 www.ajhg.org
Table 2
Allelic Association in Linkage Case-Control Series
M
ARKER
L
OCATION
(Mb)
N
O
.
OF
C
ASES
WITH
G
ENOTYPE
N
O
.
OF
C
ONTROLS
WITH
G
ENOTYPE
P
a
OR (95% CI)11 12 22 11 12 22
rs1057971 86.7 1 35 304 0 41 302 .73 .91 (.57–1.43)
rs498055 91.09 110 207 112 58 162 96 .017 1.26 (1.02–1.54)
rs4417206 91.13 45 163 145 59 139 131 .24 .87 (.70–1.09)
rs600879 102.66 285 62 6 275 56 8 1 1.01 (.72–1.43)
rs1903908 102.94 13 87 265 6 84 252 .45 1.12 (.84–1.51)
N
OTE
.—For cases, one affected sibling was genotyped from each family in the linkage sample
(Myers et al. 2002) and was compared with a set of independent controls.
a
Allelic tests are two sided except for rs498055.
Table 3
Linkage Analysis of Pedigrees Stratified by rs498055
Sample No. of Pedigrees No. Affected Peak LOD
Location
(cM)
All 343 733 3.84 68
Pedigrees with SNPs 292 624 3.18 68
Probands with A 228 488 3.37 68
Probands with G 221 471 2.38 50
Probands without A 64 136 .73 90
Probands without G 71 153 1.18 67
replication, rather than multiple testing corrections, was used
to evaluate the significance of associated SNPs.
Linkage Analysis
To determine whether rs498055 contributed to our linkage
signal on chromosome 10, we stratified families on the basis
of the presence or absence of the risk allele of rs498055 in the
proband of each family. The families used in this analysis were
the NIMH and the National Cell Repository for Alzheimer’s
Disease families from our linkage screen (Myers et al. 2002).
The analysis was performed in Mapmaker/SIBS (“All pairs,
UNWEIGHTED”). For the “proband” analysis, the (numer-
ically) first individual with the SNP genotype was identified as
the proband.
Haplotype Analysis
Several studies have shown that placing individual SNPs into
the context of a haplotype increases biological information
(Balciuniene et al. 2002; Knoblauch et al. 2002; Van Eerdew-
egh et al. 2002). Similarly, placing haplotypes into their evo-
lutionary context also increases biological information (Tem-
pleton et al. 2005). For the haplotype analysis, we used SNPs
that were typed in all three series and were located within 40
kb of rs498055. These criteria resulted in a data set of 11
SNPs in 1,159 controls and 974 cases from the WU, UK, and
UCSD case-control samples.
Haplotypes were estimated using the software PHASE (Ste-
phens et al. 2001; Stephens and Donnelly 2003). A set of 95%-
plausible haplotype trees was estimated using statistical par-
simony in the program TCS (Clement et al. 2000; Templeton
et al. 2000).
Association with LOAD was tested by tree scanning (Tem-
pleton et al. 2005), which was modified to manage case-control
data (Nowotny et al. 2005). A tree scan uses the haplotype
network to define tests that are based on each branch of the
tree. Each branch represents an a priori defined pooling of
haplotypes: haplotypes on one side of the branch are pooled
together and define an allele, whereas the haplotypes on the
other side are pooled to define a separate allele. This results
in a biallelic locus that can be tested for association with the
phenotype. A permutation-based analog of the sequential Bon-
ferroni (Westfall and Young 1993) was used to obtain nominal
and multiple-test–corrected significance values with the para-
metric P value used as the test statistic. This permutation method
takes into account the correlation structure between tests while
correcting for multiple tests.
Results
To identify genetic variation associated with LOAD on
chromosome 10, we performed a SNP-based association
study with three well-characterized LOAD case-control
series. Our strategy was to test markers in one sample
set (exploratory sample) and to follow up significant
markers in the two remaining sample sets (validation
samples). Using this paradigm, we first scanned a rela-
tively large number of gene-based putative functional
SNPs across chromosome 10, with the highest SNP den-
www.ajhg.org Grupe et al.: Association in AD 83
Figure 3 Haplotype networks. Each oval contains the haplotype identification number, the state at each locus, and the number of times
it was inferred to occur in this sample set. To simplify the presentation of the network, haplotypes that appear only once in the sample are not
shown, and selected haplotypes have been collapsed. The branch that was significant in the tree scan is denoted by the dashed line. P values
for the original and conditional analyses are also provided. Mutations at rs498055 are indicated by “TRr C”; the mutation at rs495998 is
indicated by “ARr C.”
sity in regions directly under the linkage peak reported
above. Significant markers were then genotyped in the
other two sample sets to attempt replication of the initial
association. Regions with markers showing strong asso-
ciation with the exploratory sample and replication in at
least one other sample set were then tested with addi-
tional markers. Specifically, we genotyped a total of 1,397
SNPs by allele-specific PCR in the exploratory stage (fig.
1), targeting 674 genes. From these, we genotyped 408
genes with 1 marker, 141 with 2 markers, 57 with 3
markers, 47 with 4–7 markers, and the remainder with
8 markers. The majority of exploratory markers (1,291)
were tested in the WU sample set. In the UK sample set,
105 markers were genotyped, and 1 marker was ge-
notyped in the UCSD sample set. Of the 1,397 tested
SNPs, 69 were significantly associated with LOAD in
the exploratory sample ( ). These markers wereP
! .05
scattered across the chromosome, as would be expected
because of the high probability of false-positive associ-
ations due to the large number of SNPs analyzed (fig.
1). We subsequently genotyped the 69 markers in the
two validation sample sets and found 5 that replicated
in a meta-analysis combining the two validation sample
sets (one-sided ) (table 1). One marker, rs498055,P
! .05
located in LOC439999, a gene with high homology
to RPS3A (MIM 180478), was significant ( ) inP
! .05
each of the three sample sets and was the most signifi-
cant ( ) marker in the three-sample meta-P p .00004
analysis (table 1). One other marker, rs4417206, lo-
cated in a neighboring gene ALDH18A1 (or PYCS
[MIM 138250]), was also significant in the combined
validation study ( ). Markers rs4417206 inP p .013
ALDH18A1 and rs498055 in LOC439999 are 41 kb
apart and are in strong linkage disequilibrium (LD) with
one another ( ; ).
2
D p 0.98 r p 0.43
To determine whether any of the five SNPs that rep-
licated in the meta-analysis (rs1057971, rs498055,
rs4417206, rs600879, and rs1903908) were also asso-
84 The American Journal of Human Genetics Volume 78 January 2006 www.ajhg.org
Table 4
Measures of Pairwise D
and in UK Controls
2
r
Marker Distance rs500470 rs533383 rs533343 rs11594687 rs7895441 rs495998 rs17110999 rs7906450 rs498055 rs2296690
rs500470 .00 1.00 1.00 1.00 1.00 .89 1.00 1.00 .88 .95
rs533383 5.78 .99 .99 1.00 1.00 .89 1.00 1.00 .88 .95
rs533343 .01 .74 .73 1.00 .66 .99 1.00 .92 .98 1.00
rs11594687 4.03 .07 .07 .05 1.00 1.00 1.00 1.00 1.00 1.00
rs7895441 1.40 .37 .37 .22 .03 1.00 1.00 .94 .98 1.00
rs495998 5.76 .54 .54 .48 .11 .25 1.00 1.00 .99 .77
rs17110999 2.17 .26 .27 .36 .02 .72 .18 1.00 1.00 1.00
rs7906450 5.57 .28 .28 .32 .02 .68 .19 .94 1.00 .85
rs498055 1.49 .52 .51 .47 .11 .24 .98 .18 .19 .78
rs2296690 16.03 .09 .09 .07 .69 .04 .10 .03 .02 .10
rs1804934 .20 .04 .04 .05 .00 .00 .02 .00 .00 .02 .00
rs749049 4.96 .01 .01 .00 .10 .06 .11 .19 .20 .11 .11
hDV68531050 .02 .06 .06 .02 .00 .18 .04 .01 .01 .04 .00
rs2986401 6.55 .03 .03 .04 .05 .19 .00 .33 .30 .00 .07
rs2275272 5.51 .09 .09 .07 .74 .03 .09 .02 .03 .09 .94
rs4417206 8.31 .30 .29 .21 .05 .11 .44 .08 .09 .43 .07
rs11188410 .09 .14 .13 .19 .01 .01 .10 .00 .00 .09 .01
rs10882645 7.81 .03 .03 .03 .05 .17 .00 .31 .33 .00 .06
hDV68531048 12.0 .06 .06 .02 .00 .18 .04 .01 .01 .04 .00
rs11553577 37.1 .14 .13 .19 .01 .00 .09 .00 .00 .08 .02
hCV25943811 .24 .06 .06 .02 .00 .18 .04 .01 .01 .04 .00
rs1418709 65.0 .00 .00 .02 .11 .18 .05 .12 .13 .05 .15
N
OTE
.—Measures of pairwise D
are shown above the diagonal; values are shown below the diagonal.
2
r
ciated with risk for LOAD in our original linkage study
sample (Myers et al. 2002), we genotyped the entire
series and performed two analyses. First, we used a case-
control approach by selecting one case (proband) per
family, and we matched each of them to an equal number
of unrelated controls. We chose to use a case-control
analysis rather than a discordant–sib-pair analysis be-
cause of the greater power in the case-control design and
because discordant siblings were available for only a
proportion of the cases. A one-sided x
2
test demonstrated
significant evidence of association in the case-control
sample with the same allele as in the other case-control
series for rs498055 ( ); all other SNPs failedP p .0165
to show any evidence of association (table 2). The ORs
observed in the linkage series for rs498055 were simi-
lar to those observed in the other case-control series
( ; 95% CI 1.02–1.54).OR p 1.26
Marker rs498055 was also examined in two small
series (183 cases/127 controls; 160 cases/106 controls)
of neuropathologically confirmed cases and controls. The
SNP was not associated with AD risk in these samples
( and , respectively). However, power toP p .63 P p .21
replicate our finding in these samples was low (40% and
36%, respectively; 60% power in the combined sample
sets).
To further estimate the effect of rs498055 in the link-
age sample, we performed a stratified linkage analysis
of the stage II linkage data, on the basis of the genotype
of the proband of each pedigree. We performed stratified
linkage analyses using the pedigrees in which the pro-
band had a copy of allele A and pedigrees in which the
proband had a copy of allele G (table 3). We also con-
sidered pedigrees in which the proband was a homozy-
gote for the A allele and in which the proband was a
homozygote for the G allele. The results did not show
an increase in LOD score in probands with the risk allele.
In fact, although the first two groups were roughly the
same size, the LOD score was substantially smaller in
pedigrees in which the proband had a copy of the risk
allele. This suggests that the rs498055 polymorphism
(at 91.1 Mb) may have little direct effect on the linkage
findings, which have their peak near D10S1211 (at 59.9
Mb), and that other loci contributing to disease have yet
to be found in this region.
These findings prompted us to focus further follow-
up on the region flanking these two genes. A total of 53
markers, covering a 1.49-Mb region, were typed in the
exploratory sample, and association of these SNPs was
examined. Ten of the markers resulted in a P value
!.1
in the exploratory sample, and five were significant at
(fig. 2). After genotyping these markers in theP
! .05
validation samples, rs498055 remained the only marker
that was significantly associated with LOAD in each of
the three sample sets.
We examined LD structure in this region, using ge-
notypes from the UK and the WU sample sets. We ob-
served a block of high LD extending from rs500470
to rs1418709, covering at least 190 kb of the ge-
nomic region that includes the most-significant markers,
rs498055 and rs4417206. Although the D
values among
neighboring SNPs were high, the values were generally
2
r
low (table 4). The LD structure was comparable between
cases and controls. The five significant markers with a
P value
!.05 (rs500470, rs533343, rs495998, rs498055,
www.ajhg.org Grupe et al.: Association in AD 85
rs1804934 rs749049 hDV68531050 rs2986401 rs2275272 rs4417206 rs11188410 rs10882645 hDV68531048 rs11553577 hCV25943811 rs1418709
1.00 .12 .92 .20 1.00 .98 1.00 .18 .92 .96 .92 .02
1.00 .11 .92 .20 1.00 .98 .96 .19 .92 .93 .92 .03
1.00 .04 1.00 .19 1.00 .98 1.00 .16 1.00 .97 1.00 .20
.68 1.00 .05 1.00 1.00 1.00 1.00 1.00 .05 1.00 .05 1.00
.76 .50 .95 .63 1.00 1.00 .62 .60 .95 .25 .95 .88
1.00 .37 .91 .09 .76 .99 1.00 .09 .91 .94 .91 .23
.16 1.00 1.00 .96 1.00 1.00 .26 .94 1.00 .04 1.00 .86
.27 1.00 1.00 .90 1.00 1.00 .39 .95 1.00 .17 1.00 .87
1.00 .37 .91 .08 .77 .98 .94 .08 .91 .89 .91 .23
1.00 .87 .44 1.00 1.00 1.00 1.00 .86 .44 1.00 .44 1.00
1.00 1.00 1.00 .99 .98 1.00 1.00 1.00 1.00 1.00 1.00
.02 1.00 .04 1.00 1.00 1.00 .02 1.00 .83 1.00 .41
.00 .04 1.00 .32 1.00 .93 1.00 1.00 .63 1.00 1.00
.01 .00 .02 1.00 1.00 1.00 .99 1.00 1.00 1.00 .94
.00 .13 .00 .07 1.00 1.00 1.00 .32 1.00 .32 1.00
.01 .54 .02 .23 .07 1.00 1.00 1.00 .86 1.00 .91
.00 .08 .00 .04 .01 .04 1.00 .93 1.00 .93 1.00
.01 .00 .02 .95 .07 .23 .04 1.00 1.00 1.00 .94
.00 .04 1.00 .02 .00 .02 .00 .02 .63 1.00 1.00
.00 .06 .00 .05 .01 .03 .95 .05 .00 .63 .94
.00 .04 1.00 .02 .00 .02 .00 .02 1.00 .00 1.00
.02 .16 .05 .42 .14 .41 .09 .42 .05 0 .05
and rs4417206) were all located within this block and
exhibited higher values with rs498055 than with other
2
r
neighboring SNPs. (All had with rs498055.)
2
r 1 0.43
Comparison of these results with data in the HapMap
project indicates that the block containing rs498055 ex-
tends 419 kb and contains seven genes, LOC439999,
ALDH18A1 (MIM 138250), C10orf61, ENTPD1 (MIM
601752), hCG2023951, hCG1781136, and C10orf130.
The tree-scan analysis of 11 SNPs in the region sur-
rounding rs498055 identified significant results across
many branches of the haplotype network. However, the
results of the conditional tests suggest that the associa-
tion observed at these branches is due to their location
in the network relative to a single branch. This branch
was significant in both the original ( ) and theP p .0008
conditional ( ) analyses (fig. 3). It is marked byP p .03
mutations creating the SNPs rs498055 and rs495998.
Discussion
Genetic variants in several biological candidate genes un-
der or near the chromosome 10 linkage peaks—including
mutations in CTNNA3 (MIM 607667), PLAU (MIM
191840), IDE (MIM 146680), and others—have been
reported to be associated with LOAD. However, none of
the associations in these candidate genes has been con-
sistently replicated (Alzheimer Disease Forum). Indeed,
our own studies in the case-control series used in the
present study showed no evidence of association with any
of these genes (Myers et al. 2004; Nowotny et al. 2005).
These findings suggest that the reported association may
be false, although it remains possible that the lack of
consistent replication may be due to type 1 error, genetic
heterogeneity, population stratification, and/or a small ge-
netic effect confounded by sample sizes insufficient to rep-
licate the initial reports. With the technology that was
available to us, we performed a broadly scaled and non-
biased genotyping program. This approach would inev-
itably be burdened by a requirement of multiple-testing
corrections to assess potential associations. To mitigate
this, we designed a two-step process in which we geno-
typed 1,400 SNPs in the exploratory sample set but only
69 markers in the subsequent validation sample sets. This
strategy led us to identify five SNPs, located in five genes
on chromosome 10, that are associated with LOAD. Al-
though our genotyping scan covers the entire chromo-
some 10, these significant SNPs are located relatively close
to linkage peaks identified in other studies (Bertram et al.
2000). Our analysis included 12 SNPs in IDE, 2 SNPs in
PLAU, and 32 SNPs in CTNNA3, but none was signif-
icantly associated with LOAD (Busby et al. 2004; No-
wotny et al. 2005).
The most consistently associated marker among the
five significant SNPs is rs498055, which is significant in
each of the three initially tested clinical case-control se-
ries employed here, with an allelic P value of .00004 in
the meta-analysis of the three sample sets. The replica-
tion P value of .00021 is significant even after Bonferroni
correction for 69 markers ( ), and the meta-
P p .014
analysis of these three case-control series used in the
screening paradigm is marginally significant even after
adjustment for 1,397 SNPs ( ). The linkage sam-
P p .051
ple–derived case-control series replicates these results,
whereas the smaller combined neuropathologically con-
firmed case-control sample set is not significant. The meta-
86 The American Journal of Human Genetics Volume 78 January 2006 www.ajhg.org
analysis of all six sample sets maintains that rs498055 is
significantly associated with AD risk ( ).P p .0001
The tree-scan analysis identified a single branch in
the network that is significantly associated with LOAD.
This branch is marked by mutations at rs498055 and
rs495998. Marker rs498055 is the most significant SNP
in the single-marker association tests (see table 1), and
rs495998 is in high LD with rs498055 ( ) (table
2
r p 0.98
4). This suggests that the observed effect is a mutation
on the background shared and defined by these SNPs.
It is also interesting to note that rs498055 is homopla-
sious, with mutations inferred on four different haplo-
typic backgrounds (one major and three minor haplo-
types). In some cases, the haplotype structure of a pop-
ulation allows for tests to be conducted at each branch
that is marked by a particular SNP, which provides some
evidence as to the “causal” nature of the polymorphism.
Although no association was detected at the other tran-
sitions marked by rs498055 (a result that suggests that
the SNP is not causal), the sample sizes for these tests
are too small to provide strong evidence regarding the
causality of this SNP. Inclusion of all the associated SNPs
in this region in a logistic regression analysis by use of
sequential regression (type 1) indicates that the signifi-
cance derives only from LD with rs498055; that is, no
other significant association is observed after first in-
cluding the effect of rs498055.
Marker rs498055 is located in a gene annotated as
an RPS3A homologue in the Entrez Gene database. Al-
though the function of the RPS3A homologue is un-
known, it appears that RPS3A itself is a strong bio-
logical candidate gene for AD. It has been reported
that RPS3A mediates the interaction between BCL2 (en-
coded by BCL2 [MIM 151430]) and PARP—poly(ADP-
ribose) polymerase—(PARP1 [MIM 173870]) and that
BCL2 and RPS3A together prevent apoptosis by inhib-
iting PARP activity (Hu et al. 2000; Song et al. 2002).
Thus, RPS3A is an important player in the early phase
of apoptosis, a feature observed in AD-affected brains.
However, we have been unable to detect transcripts of
the RPS3A gene by RT-PCR in RNA from multiple tis-
sues, including brain (data not shown). This may be due
to constraints in transcript-specific primer design if a
gene has multiple paralogues, as is the case with RPS3A.
Alternatively, the annotated gene may not be expressed,
and this SNP or variants that are in LD are located in
a noncoding expressed sequence, such as a microRNA.
It is also possible that this SNP, or variants that are in
LD, modulate the transcription of neighboring genes.
The SORBS1 (MIM 605264) coding sequence is located
33.7 kb downstream from this SNP and can be consid-
ered a strong biological candidate gene. It is involved in
insulin signaling and was recently reported to be up-reg-
ulated in the hippocampus of AD-affected brains com-
pared with controls (Blalock et al. 2004). ALDH18A1
(at 91.1 Mb and in tight LD with SNPs in RPS3A) en-
codes a member of the aldehyde dehydrogenase fam-
ily, which is involved in proline biosynthesis via cata-
lyzing the conversion of
L
-glutamate to
L
-glutamate 5-
phosphate.
On the basis of the results in the combined validation
sample sets, three other markers of interest were also
identified, but they are not significant in all three indi-
vidual samples. The power to replicate the original ob-
servation in the exploratory sample for these markers is
relatively low in each of the validation samples (table 1).
These markers are located in four different genes. PCGF5
(at 86.7 Mb) encodes polycomb group (PcG) ring finger
5, a component of a multimeric, chromatin-associated
PcG protein complex, which is involved in stable repres-
sion of gene activity. SORCS1 (MIM 606283) (at 102.7
Mb) encodes a type 1 receptor containing a Vps10p-
domain and a leucine-rich domain that is involved in
endocytosis and intracellular sorting. It is most abun-
dantly expressed in the brain (Hermey et al. 1999), and
its expression can be differentially affected by neuronal
activity (Hermey et al. 2004). hCG2039140 (at 102.9
Mb) is a predicted gene in the Celera Genome Assembly,
encoding a 41-aa polypeptide with no apparent homol-
ogy to any other known proteins. The potential relevance
of these genes with LOAD remains to be examined.
Moreover, it is possible that neighboring genes might
have a role in AD, since the significant SNPs we iden-
tified or variants that are in LD may affect their function.
Although the association with rs498055 was repli-
cated in the case-control series from the linkage sample,
the pedigree analyses suggest that this association did
not significantly contribute to the original linkage signal
on chromosome 10. Although the power of this analysis
is low, it suggests that there may be more than one AD
susceptibility gene on chromosome 10.
In our screen, we did not attempt to systematically
genotype chromosome 10; rather, we used an oppor-
tunistic approach to identify functionally relevant gene-
based variants that show significant association with AD
in at least two independently collected case-control sam-
ple sets. Therefore, we cannot exclude the majority of
nonsignificant chromosome 10 genes from those that
might contribute to the genetic risk of AD. This would
require high-density SNP genotyping incorporating an
LD-based approach to SNP selection in the case of the
common disease–common variant hypothesis and, ulti-
mately, deep resequencing of all genes, to exclude rare
pathogenic variants. However, the results outlined above
highlight five SNPs—particularly rs498055, which was
replicated in four independent case-control series—and
corresponding genes as likely AD risk factors on chromo-
some 10. These findings require functional experiments
to validate potential links of the genes and genetic varia-
tion to pathways related to disease mechanisms for AD.
www.ajhg.org Grupe et al.: Association in AD 87
Acknowledgments
Some of the authors are employed by Celera Diagnostics,
have personal financial interests in the company, or receive
research funding from Celera Diagnostics. Funding for this
work was partly provided by National Institutes of Health
(NIH) ADRC grants P50 AG05681 (to J.C.M.), P50 AG05131
(to L.T.), RO1 AG16208 (to A. Goate), and PO1 AG03991 (to
J.C.M.); the MRC UK (to J.W., M. Owen, S.L. M. O’Donovan,
and L. Jones); and the Alzheimer Research Trust (to J.W., M.
Owen, and S.L.). J.H. and A.M. were supported by the NIH
intramural program and by the VERUM Foundation (DIA-
DEM project). J.S.K.K. is supported by NIH training grant
T32 HG000045, and T.J.M. is supported by MICORTEX and
NIH grant GM065509. We acknowledge Mary Coats and Eliz-
abeth Grant for coordinating the Washington University ma-
terial, Mary Sundsmo for coordinating the UCSD case mate-
rial, and Pamela Moore and Dragana Turic for providing
clinical/DNA samples from MRC UK Genetic Resource for
LOAD. Many data and biomaterials were collected in three
projects that participated in the NIMH Alzheimer’s Disease
Genetics Initiative. From 1991 to 1998, the research centers,
grant numbers, and principal investigators and coinvestigators
were as follows: Massachusetts General Hospital, Boston, U01
MH46281, Marilyn S. Albert and Deborah Blacker; Johns
Hopkins University, Baltimore, U01 MH46290, Susan Bassett,
Gary A. Chase, and Marshal F. Folstein; University of Ala-
bama, Birmingham, U01 MH46373, Rodney C. P. Go and
Lindy E. Harrell. Samples for this study also came from the
National Cell Repository for Alzheimer’s Disease, which is sup-
ported by cooperative agreement NIA grant U24 AG021886.
The neuropathological series were collected largely from sev-
eral NIA–National Alzheimer’s Coordinating Center–funded
sites. Marcelle Morrison-Bogorad, Tony Phelps, and Walter
Kukull are thanked for helping to coordinate this collection.
The research centers, directors, pathologist, and technicians
involved include: NIA, Bethesda, Ruth Seemann; Johns Hop-
kins ADRC, Baltimore (NIA grant AG 05146), Juan C. Tron-
coso and Olga Pletnikova; University of California–Los An-
geles (NIA grant P50 AG16570), Harry Vinters and Justine
Pomakian; the Kathleen Price Bryan Brain Bank, Duke Uni-
versity Medical Center, Durham, NC (NIA grant AG05128,
National Institute of Neurological Disorders and Stroke grant
NS39764, NIMH grant MH60451, and support from Glaxo-
SmithKline), Christine Hulette; Stanford University, La Jolla,
Dikran Horoupian and Ahmad Salehi; New York Brain Bank,
Taub Institute, Columbia University, New York, Jean Paul
Vonsattel; Massachusetts General Hospital, Boston, E. Tessa
Hedley-Whyte and Karlotta Fitch; University of Michigan,
Ann Arbor (NIH grant P50-AG08671), Roger Albin, Lisa
Bain, and Eszter Gombosi; University of Kentucky, Lexington,
William Markesbery and Sonya Anderson, Mayo Clinic Jack-
sonville, Jacksonville, FL, Dennis W. Dickson and Natalie Tho-
mas; University of Southern California, Los Angeles, Carol A.
Miller, Jenny Tang, and Dimitri Diaz; ADRC, Washington Uni-
versity, St. Louis, Dan McKeel, John C. Morris, Eugene John-
son Jr., Virginia Buckles, and Deborah Carter; University of
Washington, Seattle, Thomas Montine and Aimee Schantz.
A.J.M. is a resident research associate of the National Academy
of Sciences (U.S.A.).
Web Resources
The URLs for data presented herein are as follows:
Alzheimer Disease Forum, http://www.alzforum.org/res/com/gen/
alzgene/chromo.asp?cp10
Entrez Gene, http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?dbpgene
Online Mendelian Inheritance in Man (OMIM), http://www.ncbi.nlm
.nih.gov/Omim/ (for AD, APP, PSEN1, PSEN2, LOAD, APOE,
GAPD, RPS3A, PYCS, ALDH18A1, ENTPD1, CTNNA3, PLAU,
IDE, BCL2, PARP1, SORBS1, and SORCS1)
References
Agresti A (1990) Categorical data analysis. John Wiley & Sons, New
York
Balciuniene J, Emilsson L, Oreland L, Pettersson U, Jazin E (2002)
Investigation of the functional effect of monoamine oxidase poly-
morphisms in human brain. Hum Genet 110:1–7
Berg L, McKeel DW Jr, Miller JP, Storandt M, Rubin EH, Morris JC,
Baty J, Coats M, Norton J, Goate AM, Price JL, Gearing M, Mirra
SS, Saunders AM (1998) Clinicopathologic studies in cognitively
healthy aging and Alzheimer’s disease: relation of histologic markers
to dementia severity, age, sex, and apolipoprotein E genotype. Arch
Neurol 55:326–335
Bertram L, Blacker D, Mullin K, Keeney D, Jones J, Basu S, Yhu S,
McInnis MG, Go RC, Vekrellis K, Selkoe DJ, Saunders AJ, Tanzi
RE (2000) Evidence for genetic linkage of Alzheimer’s disease to
chromosome 10q. Science 290:2302–2303
Blacker D, Bertram L, Saunders AJ, Moscarillo TJ, Albert MS, Wiener
H, Perry RT, Collins JS, Harrell LE, Go RC, Mahoney A, Beaty T,
Fallin MD, Avramopoulos D, Chase GA, Folstein MF, McInnis MG,
Bassett SS, Doheny KJ, Pugh EW, Tanzi RE (2003) Results of a high-
resolution genome screen of 437 Alzheimer’s disease families. Hum
Mol Genet 12:23–32
Blalock EM, Geddes JW, Chen KC, Porter NM, Markesbery WR,
Landfield PW (2004) Incipient Alzheimer’s disease: microarray cor-
relation analyses reveal major transcriptional and tumor suppressor
responses. Proc Natl Acad Sci USA 101:2173–2178
Busby V, Goossens S, Nowotny P, Hamilton G, Smemo S, Harold D,
Turic D, et al (2004) a-
T
-catenin is expressed in human brain and
interacts with the Wnt signaling pathway but is not responsible for
linkage to chromosome 10 in Alzheimer’s disease. Neuromolecular
Med 5:133–146
Clement M, Posada D, Crandall KA (2000) TCS: a computer program
to estimate gene genealogies. Mol Ecol 9:1657–1659
Corder EH, Saunders AM, Strittmatter WJ, Schmechel DE, Gaskell
PC, Small GW, Roses AD, Haines JL, Pericak-Vance MA (1993)
Gene dose of apolipoprotein E type 4 allele and the risk of Alzhei-
mer’s disease in late onset families. Science 261:921–923
Ertekin-Taner N, Graff-Radford N, Younkin LH, Eckman C, Baker
M, Adamson J, Ronald J, Blangero J, Hutton M, Younkin SG (2000)
Linkage of plasma Ab42 to a quantitative locus on chromosome 10
in late-onset Alzheimer’s disease pedigrees. Science 290:2303–2304
Farrer LA, Bowirrat A, Friedland RP, Waraska K, Korczyn AD, Baldwin
CT (2003) Identification of multiple loci for Alzheimer disease in a
consanguineous Israeli-Arab community. Hum Mol Genet 12:415
422
Germer S, Holland MJ, Higuchi R (2000) High-throughput SNP allele-
frequency determination in pooled DNA samples by kinetic PCR.
Genome Res 10:258–266
Goate A, Chartier-Harlin M-C, Mullan M, Brown J, Crawford F, Fi-
dani L, Giuffra L, Haynes A, Irving N, James L, Mant R, Newton
P, Rooke K, Roques P, Talbot C, Pericak-Vance M, Roses A, Wil-
liamson R, Rossor M, Owen M, Hardy J (1991) Segregation of a
88 The American Journal of Human Genetics Volume 78 January 2006 www.ajhg.org
missense mutation in the amyloid precursor protein gene with familial
Alzheimer’s disease. Nature 349:704–706
Hermey G, Plath N, Hubner CA, Kuhl D, Schaller HC, Hermans-
Borgmeyer I (2004) The three Sorcs genes are differentially expressed
and regulated by synaptic activity. J Neurochem 88:1470–1476
Hermey G, Riedel IB, Hampe W, Schaller HC, Hermans-Borgmeyer I
(1999) Identification and characterization of SorCS, a third member
of a novel receptor family. Biochem Biophys Res Commun 266:347–
351
Hu ZB, Minden MD, McCulloch EA, Stahl J (2000) Regulation of
drug sensitivity by ribosomal protein S3a. Blood 95:1047–1055
Kehoe P, Wavrant-De Vrieze F, Crook R, Wu WS, Holmans P, Fenton
I, Spurlock G, Norton N, Williams H, Williams N, Lovestone S,
Perez-Tur J, Hutton M, Chartier-Harlin MC, Shears S, Roehl K,
Booth J, Van Voorst W, Ramic D, Williams J, Goate A, Hardy J,
Owen MJ (1999) A full genome scan for late onset Alzheimer’s
disease. Hum Mol Genet 8:237–245
Knoblauch H, Bauerfeind A, Krahenbuhl C, Daury A, Rohde K, Be-
janin S, Essioux L, Schuster H, Luft FC, Reich JG (2002) Common
haplotypes in five genes influence genetic variance of LDL and HDL
cholesterol in the general population. Hum Mol Genet 11:1477–
1485
Levy-Lahad E, Wasco W, Poorkaj P, Romano DM, Oshima J, Pettingell
WH, Yu CE, Jondro PD, Schmidt SD, Wang K, Crowley A, Fu Y-
H, Guenette S, Galas D, Nemens E, Wijsman E, Bird T, Schellenberg
G, Tanzi R (1995) Candidate gene for the chromosome 1 familial
Alzheimer’s disease locus. Science 269:973–977
Li Y, Hollingworth P, Moore P, Foy C, Archer N, Powell J, Nowotny
P, Holmans P, O’Donovan M, Tacey K, Doil L, van Luchene R,
Garcia V, Rowland C, Lau K, Cantanese J, Sninsky J, Hardy J, Thal
L, Morris JC, Goate A, Lovestone S, Owen M, Williams J, Grupe
A (2005) Genetic association of the APP binding protein 2 gene
(APBB2) with late onset Alzheimer disease. Hum Mutat 25:270–
277
Li Y, Nowotny P, Holmans P, Smemo S, Kauwe JS, Hinrichs AL, Tacey
K, et al (2004) Association of late-onset Alzheimer’s disease with
genetic variation in multiple members of the GAPD gene family.
Proc Natl Acad Sci USA 101:15688–15693
McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan
EM (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
Myers A, Holmans P, Marshall H, Kwon J, Meyer D, Ramic D, Shears
S, Booth J, DeVrieze FW, Crook R, Hamshere M, Abraham R, Tun-
stall N, Rice F, Carty S, Lillystone S, Kehoe P, Rudrasingham V,
Jones L, Lovestone S, Perez-Tur J, Williams J, Owen MJ, Hardy J,
Goate AM (2000) Susceptibility locus for Alzheimer’s disease on
chromosome 10. Science 290:2304–2305
Myers A, Wavrant De-Vrieze F, Holmans P, Hamshere M, Crook R,
Compton D, Marshall H, et al (2002) Full genome screen for Alz-
heimer disease: stage II analysis. Am J Med Genet 114:235–244
Myers AJ, Marshall H, Holmans P, Compton D, Crook RJ, Mander
AP, Nowotny P, Smemo S, Dunstan M, Jehu L, Wang JC, Hamshere
M, Morris JC, Norton J, Chakraventy S, Tunstall N, Lovestone S,
Petersen R, O’Donovan M, Jones L, Williams J, Owen MJ, Hardy
J, Goate A (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
Nowotny P, Hinrichs AL, Smemo S, Kauwe J, Maxwell T, Holmans P,
Hamshere M, et al (2005) Association studies between risk for late-
onset Alzheimer’s disease and variants in insulin degrading enzyme.
Am J Med Genet B Neuropsychiatr Genet 136:62–68
Pastor P, Goate AM (2004) Molecular genetics of Alzheimer’s disease.
Curr Psychiatry Rep 6:125–133
Pericak-Vance MA, Bass MP, Yamaoka LH, Gaskell PC, Scott WK,
Terwedow HA, Menold MM, Conneally PM, Small GW, Vance JM,
Saunders AM, Roses AD, Haines JL (1997) Complete genomic
screen in late-onset familial Alzheimer disease: evidence for a new
locus on chromosome 12. JAMA 278:1237–1241
Pericak-Vance MA, Grubber J, Bailey LR, Hedges D, West S, Santoro
L, Kemmerer B, Hall JL, Saunders AM, Roses AD, Small GW, Scott
WK, Conneally PM, Vance JM, Haines JL (2000) Identification of
novel genes in late-onset Alzheimer’s disease. Exp Gerontol 35:
1343–1352
Rogaeva E, Premkumar S, Song Y, Sorbi S, Brindle N, Paterson A,
Duara R, Levesque G, Yu G, Nishimura M, Ikeda M, O’Toole C,
Kawarai T, Jorge R, Vilarino D, Bruni AC, Farrer LA, St George-
Hyslop PH (1998) Evidence for an Alzheimer disease susceptibility
locus on chromosome 12 and for further locus heterogeneity. JAMA
280:614–618
Sherrington R, Rogaev EI, Liang Y, Rogaeva EA, Levesque
G, Ikeda M, Chi H, et al (1995) Cloning of a gene bearing missense
mutations in early-onset familial Alzheimer’s disease. Nature 375:
754–760
Song D, Sakamoto S, Taniguchi T (2002) Inhibition of poly(ADP-
ribose) polymerase activity by Bcl-2 in association with the ribo-
somal protein S3a. Biochemistry 41:929–934
Stephens M, Donnelly P (2003) A comparison of Bayesian methods
for haplotype reconstruction from population genotype data. Am J
Hum Genet 73:1162–1169
Stephens M, Smith NJ, Donnelly P (2001) A new statistical method
for haplotype reconstruction from population data. Am J Hum Ge-
net 68:978–989
Templeton AR, Maxwell T, Posada D, Stengard JH, Boerwinkle E, Sing
CF (2005) Tree scanning: a method for using haplotype trees in
phenotype/genotype association studies. Genetics 169:441–453
Templeton AR, Weiss KM, Nickerson DA, Boerwinkle E, Sing CF
(2000) Cladistic structure within the human lipoprotein lipase gene
and its implications for phenotypic association studies. Genetics 156:
1259–1275
Van Eerdewegh P, Little RD, Dupuis J, Del Mastro RG, Falls K, Simon
J, Torrey D, et al (2002) Association of the ADAM33 gene with
asthma and bronchial hyperresponsiveness. Nature 418:426–430
Westfall P, Young S (1993) Resampling-based multiple testing. John
Wiley & Sons, New York
    • "There are many reported pleiotropic 25 effects of statins. For example, chronic statin treatment 26 can enhance memory (e.g., atorvastatin or simvastatin, 27 (Grupe et al., 2006; Li et al., 2006; Lu et al., 2007), 28 decrease inflammatory cytokine production (Balduini 29 et al., 2003), improve cerebral blood flow to a site of injury 30 (Chen et al., 2003) and reduce deficits in long-term poten- 31 tiation in a mouse model of Alzheimer's disease (AD) 32 (Metais et al., 2014). Moreover statins can enhance neuro- 33 genesis in the dentate gyrus (Chen et al., 2003; Lu et al., 34 2007) in addition to promoting angiogenesis and neurite 35 outgrowth (Pooler et al., 2006). "
    [Show abstract] [Hide abstract] ABSTRACT: Simvastatin is an HMG-CoA reductase inhibitor commonly used in the clinic to treat hypercholesterolemia. In addition, simvastatin has been shown to cross the blood brain barrier and pleiotropic effects of simvastatin have been reported including anti-inflammatory properties, enhancement of neurite outgrowth, and memory enhancement properties. However, little has been reported on the effects of simvastatin on basal synaptic transmission and neuronal excitability. Here we report that simvastatin increases the fEPSP, the NMDA receptor-mediated fEPSP using extracellular recordings in the dendritic region of the CA1 of hippocampal slices taken from 8 week old C57Black6J mice. In addition, we found that simvastatin perfusion causes a change in the input/output curve and a decrease of the paired pulse facilitation ratio, indicating respectively an increase of the neuronal excitability and neurotransmitter release. We have also observed that acute application of simvastatin increased the amplitude of the compound action potential in the CA1 region. Notably, using LY294002, we have demonstrated that this effect was PI3K dependent and was occluded if the animals had previously received a diet supplemented with simvastatin. We have finally shown that the simvastatin mediated increase of the compound action potential amplitude was also present in hippocampal slices from aged mice. Copyright © 2015. Published by Elsevier Ltd.
    Full-text · Article · Feb 2015
    • "GABRE is related to migraine suscept- ibility [49]. HKDC1 is related to Alzheimer disease [50]. And LRRC1 is DNA repair related [51]. "
    [Show abstract] [Hide abstract] ABSTRACT: Hepatocellular carcinoma (HCC) is one of the most highly malignant and lethal cancers of the world. Its pathogenesis has been reported to be multi-factorial, and the molecular carcinogenesis of HCC can not be attributed to just a few individual genes. Based on the microRNA and mRNA expression profiling of normal liver tissues, pericancerous hepatocellular tissues and hepatocellular carcinoma tissues, we attempted to find prognosis related gene sets for HCC patients. We identified differentially expressed genes (DEG) from three comparisons: Cancer/Normal, Cancer/Pericancerous and Pericancerous/Normal. GSEA (gene set enrichment analysis) were performed. Based on the enriched gene sets of GO terms, pathways and transcription factor targets, it was found that the genome instability and cell proliferation increased while the metabolism and differentiation decreased in HCC tissues. The expression profile of DEGs in each enriched gene set was used to correlate to the postoperative survival time of HCC patients. Nine gene sets were found to prognostic correlation. Furthermore, after substituting DEG-targeting-microRNA for DEG members of each gene set, two gene sets with the microRNA expression profiles were obtained that had prognostic potential. The malignancy of HCC could be represented by gene sets, and pericancerous liver exhibits important characteristics of liver cancer. The expression level of gene sets not only in HCC but also in the pericancerous liver showed potential for prognosis implying an option for HCC prognosis at an early stage. Additionally, the gene-targeting-microRNA expression profiles also showed prognostic potential, demonstrating that the multi-factorial molecular pathogenesis of HCC is contributed by various genes and microRNAs.
    Full-text · Article · Jan 2014
    Lin WeiLin WeiBaofeng LianBaofeng LianYuannv ZhangYuannv Zhang+1more author...[...]
    • "To date, it has been demonstrated that the overexpression of SORCS1 reduced γ-secretase activity and Aβ levels in other studies (Lane et al., 2010; Reitz et al., 2011). On the other hand, one genome-wide association study (GWAS) and chromosome10-specific association studies have identified SORCS1 as a candidate gene for AD, but different single-nucleotide polymorphisms (SNPs) in SORCS1 were reported to be associated with AD in these studies (Grupe et al., 2006; Li et al., 2008; Liang et al., 2009). Recently in a large-scale case–control study including six independent datasets, Reitz and colleagues investigated the genetic association between the 16 SNPs mainly reported by the previous studies and AD [3] . "
    [Show abstract] [Hide abstract] ABSTRACT: Sortilin-related VPS domain containing receptor 1 (SORCS1), is located on chromosome 10q23.3, a chromosomal region of interest in Alzheimer's disease (AD) defined by many genome-wide and chromosome10-specific studies. Recently, three intronic variants (rs12571141, rs17277986 and rs6584777) within SORCS1 were reported to be associated with AD in Caucasian. In order to assess the involvement of the SORCS1 polymorphisms in the progression of late-onset AD (LOAD), we conducted an independent replication study in 1198 unrelated Northern Han Chinese subjects comprising 598 LOAD patients and 600 healthy controls matched for gender and age. The results revealed no significant differences in the distributions of genotype or allele between LOAD and control groups in the total sample. However, when these data were stratified by the Apolipoprotein E (APOE) ε4 status, we observed significant differences in the genotypes and allele frequencies (rs12571141: P=0.001, rs17277986: P=0.005, rs6584777: P=0.023) in APOE ε4 allele carriers. Moreover, the association was further demonstrated in logistic regression analysis (rs12571141: P=0.002, OR=0.424; rs17277986: P=0.004, OR=0.447; rs6584777: P=0.019, OR=0.523) and haplotype analysis (GCC: P=0.002, ATT: P=0.002, ACC: P=0.025) in this subset. Our data suggested that SORCS1 was in interaction with APOE in the development of LOAD in a Northern Han Chinese population.
    Article · Feb 2012
    Hui-Fu WangHui-Fu WangJin-Tai YuJin-Tai YuWei ZhangWei Zhang+1more author...[...]
Show more