Tải bản đầy đủ

Screening of early and late onset alzheimer’s disease genetic risk factors in a cohort of dementia patients from liguria, italy

Send Orders for Reprints to reprints@benthamscience.ae
Current Alzheimer Research, 2015, 12, 802-812

802

Screening of Early and Late Onset Alzheimer’s Disease Genetic Risk
Factors in a Cohort of Dementia Patients from Liguria, Italy
Raffaele Ferrari1, Michela Ferrara2, Anwar Alinani1, Roger Brian Sutton3, Francesco Famà2,
Agnese Picco2, Guido Rodriguez2, Flavio Nobili2 and Parastoo Momeni1,*
1

Texas Tech University, Health Sciences Center, Department of Internal Medicine, 3601 4 th Street, Lubbock, TX, 79430,
USA; 2University of Genoa, Clinical Neurology, Department of Neurosciences (DINOGMI), Largo Paolo Daneo 3,
16132, Genoa, Italy; 3Texas Tech University, Health Sciences Center, Department of Cell Physiology and Molecular
Biophysics, 3601 4th Street, Lubbock, TX, 79430, USA
Abstract: Cohorts from a defined geographical area enable ad hoc genotype-phenotype correlation studies providing
novel and unique insight into disease. We analysed genetic risk factors associated with early and late onset Alzheimer’s
disease (EOAD and LOAD) in a population from Liguria (northern Italy), as part of an ongoing longitudinal study. We
screened 37 AD, 8 mild cognitive impairment (MCI), 3 AD and CVD (cerebrovascular disease), 3 MCI and CVD, 8 frontotemporal dementia (FTD) and 2 progressive supranuclear palsy (PSP) patients, and 28 normal controls (NCs). We sequenced PSEN1, PSEN2 and APP (EOAD risk factors), as well as MAPT, GRN and TARDBP for all cases and NCs, and
analysed the APOE, CLU, CR1 and PICALM genotypes as well as the MAPT and ACE haplotypes (LOAD risk factors)
for the AD (n = 37) and AD + MCI (n = 45) cases and NCs (n = 28). We identified variants in PSEN1, PSEN2 and

TARDBP across a range of phenotypes (AD, AD and CVD, FTD and PSP), suggesting that screening of all known candidate genes of Alzheimer’s and non-Alzheimer’s forms of dementias in all dementia cases might be warranted. The analysis of the LOAD risk factors revealed no association with AD or AD + MCI status after Bonferroni correction. Lack of association with APOE is supported by previous studies in the Italian population. Our data also evidenced: 1) a potentially
protective haplotype at the PSEN2 locus; 2) a nominal association with the GWAS-risk allele A for rs3818361 in CR1
and; 3) a threefold prevalenceof AD in the femalepopulation compared to men. Our results will need to be further assessed and confirmed in larger cohorts from this area.

Keywords: EOAD, gender, genotyping, LOAD, population specificity, risk factors, sequencing.
1. INTRODUCTION
Recent advancements in the biomedical fields and overall
improved life standards in the Western world have accounted
for gradual population growth leading to an increase of disorders in the elderly, including dementias [1]. In 2010 ~36
million people over 60 years of age were estimated estimated
to have dementia, worldwide [1]. Alzheimer’s disease (AD)
is the most prevalent form of dementia, currently affecting
5.2 million people in the United States (US) [2, 3]; in the
European Union (EU) AD is projected to reach 16 million
cases by 2040 [4]. The incidence of AD is expected to almost triple in the next decades reaching 106 million affected
individuals by 2050, worldwide [5].
AD is subdivided in two major groups based on the age
of onset. Early onset AD (EOAD; age of onset <65) occurs
in ≤ 5% of cases, ~1% of which is associated with mutations
in the amyloid precursor protein (APP), presenilin-1
(PSEN1) or presenilin-2 (PSEN2) genes [6]. Late onset AD
(LOAD; age of onset ≥65) occurs in ≥ 95% of cases and is
*Address correspondence to this author at the Department of Internal Medicine 4C179/4B134-135, Texas Tech University Health Sciences Center,
3601 4th St. STOP 9410, Lubbock Texas 79430, USA; Tel: 806-743-6843;
Fax: 806-743-3148; E-mail: parastoo.momeni@ttuhsc.edu

1875-5828/15 $58.00+.00

complex and multifactorial [7]. The apolipoprotein E
(APOE) is currently the main known genetic risk factor for
LOAD [8]; the E4 allele appears to be associated with increased risk of developing AD [9] and the APOE locus has
consistently been found to be associated with AD in genome
wide association studies (GWAS). The list of loci/genes associated with AD has been and is being enriched by recent
GWAS [10-13]: to date, the most relevant and replicated loci
include phosphatidylinositol binding clathrin assembly protein (PICALM), clusterin (CLU) and complement receptor 1
(CR1) [11, 12]. These genetic factors have a small effect size
but provide a basis for further and focused investigation of
the patho-mechanisms of AD [14]. Interestingly, angiotensin
I converting enzyme (ACE) [15] and the microtubule associated protein (MAPT) haplotypes [16, 17] are factors suggested to influence risk of developing AD supporting the
idea that haplotype substructures might drive or influence

disease.
Ageing is the main risk factor for AD [18]; in addition,
gender [3, 18] as well as genetic (see above), epigenetic [19,
20], environmental factors such as intelligence quotient (IQ),
education and physical activity have been recognised for
their implication in disease development [4]. However, the
lack of understanding of disease aetiology and the scarcity of

© 2015 Bentham Science Publishers


Screening of EOAD and LOAD Risk Factors

reliable biomarkers to diagnose at an early stage and monitor
disease reflect the critical fact that, to date, no successful
preventive or therapeutic measures for AD are available [15].
Subtle memory impairments, as observed in mild cognitive
impairment (MCI), are among the first symptoms of AD and
within the complex MCI syndrome, both an amnestic
(aMCI) and non-amnestic (naMCI) type have been described
[21]. The diagnosis of aMCI reflects objective memory impairment in the absence of dementia [22], thus it had been
hypothesised that a diagnosis of aMCI could be useful for
detecting AD in its early stages [23, 24].
Our long-term goal is to perform a longitudinal study in a
population recruited in Liguria, northern Italy, from the metropolitan area of Genoa with the aim of testing the above
mentioned hypothesis. Our current study is performed on an
initial sample set that has been recently recruited and is part
of a structured collaborative effort that intends to interactively store, manage and correlate in depth assessments of
MCI and AD cases combining clinical, neuropsychological,
imaging and genetic data [25]. Furthermore, study cohorts of
this kind better enable evaluation of the genetic background
against the exposure to common environmental factors with
the potential to better characterise genotype-phenotype correlations, providing novel and unique insight into disease.
Herein we present the results of the screening of known
genes for EOAD (PSEN1, PSEN2 and APP) and the risk
factors associated with LOAD such as the APOE, CLU, CR1
and PICALM genotypes, and the MAPT and ACE haplotypes.
We evaluated the presence and distribution of potential
pathogenic variants in the candidate genes in the entire study
population, whereas we analysed association with disease of
the LOAD risk factors (APOE, CLU, CR1 and PICALM) and
risk haplotypes (MAPT and ACE in first instance) in the AD
and AD + MCI cases only. In addition, we also sequenced
the MAPT, progranulin (GRN) and the TAR DNA binding
protein 43 (TARDBP) genes in the entire cohort.
2. MATERIALS AND METHODS
2.1. Study Subjects
Patients with cognitive complaints and objective cognitive impairment as established by means of an extended neuropsychological battery were evaluated at the University of
Genoa. At the end of a complete clinical evaluation including neuroimaging, a diagnosis was made and patients started
appropriate treatment and clinical follow-up. The majority of
patients were diagnosed with either AD or MCI; a minority
of cases had comorbidities such as cerebrovascular disease
(CVD), whereas a small proportion of patients were diagnosed with frontotemporal dementia (FTD) or progressive
supranuclear palsy (PSP) (Supplementary Table 1a-b). To
assess the diagnosis of MCI and/or AD, the following was
implemented: Mini Mental State Examination (MMSE) [26],
Clinical Dementia Rating scale (CDR), Geriatric Depression
Scale (GDS-15), Neuropsychiatric Inventory (NPI), questionnaires for the Activities of Daily Living (ADL) [27] and
Instrumental ADL (IADL) [28] scales, biochemical screening, drug use anamnesis and neuropsychological testing.
FTD and PSP cases were diagnosed based on the current
diagnostic criteria [29-31]. Magnetic resonance imaging
(MRI) and 18F-fluodexyglucose positron emission tomogra-

Current Alzheimer Research, 2015, Vol. 12, No. 8

803

phy (FDG-PET) scans were performed on most of the enrolled cases. Written consent to participate in this research
project was obtained from each participant or their next of
kin and the institutional review board (IRB) approved the
study at each collaborative site. Age and ethnicity matched
controls were also collected and included in the study.
2.2. Genetic Screening
Genomic DNA for the genetic screening was extracted
from blood following standard methods as recommended by
manufacturer (Promega, Madison, WI, USA).We sequenced
all the coding exons and flanking intronic regions of the
known EOAD genes such as PSEN1 (exons 4-13), PSEN2
(exons 4-13), APP (exons 16-17). Then we sequenced the
MAPT gene (exons 1, 9-13) and determined the MAPT
haplotype by PCR; in addition, we sequenced the GRN (exons 1-12) and TARDBP (exons 1 and 6) genes. Primers for
PCR amplification were designed using Primer3
(http://frodo.wi.mit.edu/primer3/). Purified PCR amplicons
were primarily sequenced in one direction and in both directions to confirm any identified variant. Sanger sequencing
was carried out with the Big Dye Terminator kit (ABI, Foster City, CA, USA) following standard protocol as recommended by manufacturer, and run on a 3730 DNA Analyzer
(ABI), followed by analysis with Sequencher 4.9 software
(Gene Codes Corporation, Ann Arbor, MI, USA). The sequencing experiments were carried out in 37 AD, 8 MCI, 3
AD and CVD (cerebrovascular disease), 3 MCI and CVD, 8
FTD and 2 PSP patients and 28 age and sex matched normal
controls collected from the same metropolitan area.We calculated the minor allele frequencies (MAF) of each variant in
our study population and compared them with those of the
general population with European ancestry available from
1000 Genomes (http://browser.1000genomes.org/index.html;
accessed
in
October
2014)
and
dbSNP
(http://www.ncbi.nlm.nih.gov/SNP/; accessed in October
2014).The genotypes were determined through allelic discrimination experiments using the SNP genotyping assay
and the 7900HT Fast Real-Time PCR System (ABI) for
rs429358 and rs7412 (APOE), rs11136000 (CLU),
rs3818361 (CR1), rs3851179 (PICALM) and rs4291, rs4295,
rs4311, rs4329, rs4343 and rs4353 (ACE) for the entire cohort and the controls.
2.3. Haplotype Analysis
Haplotype structures were analysed by means of the online free tool Haploview [32]. Specifically, we compared the
haplotypes of the cases diagnosed with AD only (n=37) or
MCI + AD (n=45) vs. normal controls (NC; n=28) for the
following loci: PSEN1 (rs63750592, rs3025786, rs165932
and rs17125721), PSEN2 (rs11405, rs6759, rs1295643,
rs1295644,
rs59683545,
rs58973334,
rs1046240,
rs63750197, rs149734051, rs200754713, rs61730652 and
rs2855562), MAPT (rs17650901, rs62063850, rs1052553,
rs17652121, rs11568305, rs3744460, rs75534191 and
rs66499584), GRN (rs60100877, rs9897526, rs25646,
rs850713,
rs72824736,
rs63750541,
rs148213321,
rs140298583, rs63750412 and rs5848), TARDBP (rs968545,
rs11121679, rs11121680, rs4133584, rs11121681 and
rs80356723) and ACE (rs4291, rs4295, rs4311, rs4329,
rs4343 and rs4353).


804 Current Alzheimer Research, 2015, Vol. 12, No. 8

2.4. In silico Assessment for Effects on Transcription and
Protein Structure/Function
Both non-coding and coding variants were investigated
for potential effects on transcription in brain using online
freely available databases such as the GTEx portal
(www.gtexportal.org/; accessed in October 2014) and the
braineac (http://www.braineac.org/; accessed in October
2014). Effects of any variant on transcription were considered significant when p-value was ≤ 1x10-5, as recommended
by the databases’ curators. Conversely, coding variants leading to missense transitions were examined in silico for potential effects on protein structure and/or function using multiple online freely available tools (accessed in October 2014)
such as: PolyPhen-2 [33], SIFT ([34] and through Ensembl
[http://www.ensembl.org/index.html]), PANTHER subPSEC
[35], SNPs3D [36], and UMD HTC [37]. We also evaluated
the 3D structures of the proteins to assess potential biochemical effects of the missense changes. Fragments of presenilin1, progranulin, and TARDBP are present in the PBD
(protein data bank), but most of the mutations of interest did
not fall within the available 3D structures. Only nuclear
magnetic resonance (NMR) structure data for the presenilin1 CTF (C-terminal fragment) was available and allowed
analysis of the missense change Glu318Gly in PSEN1.
2.5. Statistical Analysis
We assessed the frequency and distribution of the three
MAPT haplotype combinations (H1/H1, H1/H2 and H2/H2)
and the H1 and H2 alleles in either AD only (n=37) or MCI
+ AD (n=45) vs. normal controls (NC; n=28). Similarly, we
evaluated the APOE, CLU, CR1 and PICALM genotyping
results assessing the frequency and distribution of all genotypes (2/2, 2/3, 2/4, 3/3, 3/4 and 4/4) and alleles (APOE*2,
APOE*3 and APOE*4) for APOE or the risk alleles for CLU
(C), CR1 (A), and PICALM (C) in either AD only (n=37) or
MCI + AD (n=45) vs. NCs (n=28). Association with disease
status was evaluated in either AD only (n=37) or MCI + AD
(n=45) vs. NCs (n=28) using the Fisher’s exact and/or the
Chi2 test calculating p-values with significance threshold set
at p<0.05 (http://vassarstats.net/odds2x2.html; accessed in
October 2014). Bonferroni correction was implemented
when applicable or needed.
3. RESULTS
The results of the genetic screening and the effects on
expression and proteins (entire section 3.1) are summarised
in Supplementary Table 2a-b.
3.1. Sequencing Analyses of the Entire Cohort
3.1.1. APP
Sequencing of the APP gene did not reveal any variant in
any of the cases nor controls.
3.1.2. PSEN1
Screening of PSEN1 revealed two coding and two noncoding variants. One patient diagnosed with AD (RP6609)
carried the coding variant rs63750592 that causes the missense change Arg35Gln (MAF: 0.008). This variant was predicted to be benign by all in silico prediction-tools (100% of

Ferrari et al.

accordance) and had been previously reported as nonpathogenic [38, 39]; however, it was not found in our controls and it is almost absent from the general population
(MAF: 0.001). RP6609 was diagnosed while 78 years of age
and is currently 90 with MMSE = 17/30 (recently evaluated),
thus showing slow progression of disease. There is no history of dementia in the family of this index patient. Three
patients diagnosed respectively with AD (EM0408), FTD
(GA1909) and PSP (AC3109) carried a base transition leading to the missense change Glu318Gly (rs17125721). This
variant has been previously described as non-pathogenic [40,
41] or non-fully penetrant [38, 42], and it was not identified
in our controls. Its MAF was 0.024 in our cases whereas it is
currently reported as 0.009 (1000G) or varying between
0.017-0.026 (dbSNP) in the general population. In silico
assessment revealed this missense change being benign
(100% of accordance). From the NMR structure of the presenilin-1 CTF (C-terminal fragment) it is evident that
Glu318 occurs on a flexible, poly-acidic loop that connects a
single α-helix to a five helix bundle (Supplementary Fig. 1);
this mutation would disrupt the negative charge in this loop,
as well as introduce additional flexibility. This variant was
identified in three individuals with heterogeneous clinical
manifestations (AD, FTD and PSP). EM0408 was diagnosed
early (66 years of age) with AD and presented early seizures
after the occurrence of cognitive symptoms. There is a history of dementia in this family (mother diagnosed with nonspecified dementia at age 70 and deceased at age 78).
GA1909 is a sporadic FTD patient with behavioural variant
syndrome (bvFTD), age of onset 73 and comorbid condition
(diabetes mellitus). AC3109 is a patient diagnosed with PSP
with onset age 71 and family history of dementia from the
maternal side (who died at the age of 87). Provided that
analysis of larger control cohorts is warranted the clinical
significance and pathogenicity of this variant remains for the
most part uncertain [43]. The non-coding variants rs3025786
and rs165932 were similarly identified in cases and controls,
and their MAF resembled that reported in 1000G and dbSNP
for the general population. No significant expression quantitative trait loci (eQTL) were evident at this locus.
3.1.3. PSEN2
Screening of PSEN2 revealed eight coding and five noncoding variants. Five of eight coding changes were synonymous (rs11405, rs6759, rs1046240, rs61730652 and one
novel variant [G>A on chromosome 1; position: 226885631,
GRCh38]). Three (rs11405, rs6759 and rs1046240) were
evenly distributed among our cases and controls, whereas
two were exclusive to AD cases: FS4709 (novel bp transition; MAF: 0.008) and VB4309 (rs61730652; MAF: 0.008).
The known variants (rs11405, rs6759, rs1046240 and
rs61730652) in our cohort had similar MAF compared to the
general population. Three coding variants resulted in nonsynonymous changes: rs58973334 (Arg62His) was predicted
to be benign (100% of accordance) and was found in 1 AD
case (GB8009; MAF: 0.008) and one control (GA4009;
MAF: 0.018). GB8009 is an AD patient with age of onset 83
whereas GA4009, age at visit 61 and current age 68, is currently confirmed as a healthy control despite signs of hypertension and hypercholesterolemia. Previous reports classify
the pathogenicity of this variant as unclear [44, 45]; however, our findings might currently suggest this being a non-


Screening of EOAD and LOAD Risk Factors

pathogenic missense change. Rs63750197 (Ser130Leu) was
predicted to be possibly pathogenic (80% of accordance); it
was identified in one AD + CVD case (GR4409; MAF:
0.008) and none of our controls. GR4409 has age of onset 73
and shows slow progression of disease. This variant is almost absent from the general population (MAF: 0.001
[1000G and dbSNP]) and was previously shown to cosegregate with disease (EOAD) [46]. Rs200754713
(Tyr231Cys) was predicted to be probably pathogenic (100%
of accordance) but we identified it exclusively in one control
(MRP5309; MAF: 0.018). MRP5309, age at visit 67 and
current age 76, has been described presenting chronic minor
depression with recurrent relapses of mild entity and presence of moderate subcortical vascular disease due to hypertension (detected through imaging data). The Tyr231Cys
variant was previously described as pathogenic in an Italian
behavioural variant FTD (bvFTD) patient [47], nevertheless
our results now potentially challenge that report. Both controls (GA4009 and MRP5309) will need to be followed longitudinally to confirm their disease-free status. All the noncoding variants (rs1295643, rs1295644, rs59683545,
rs149734051 and rs2855562) were evenly identified in cases
and controls, and their MAF resembled that reported in
1000G and dbSNP for the general population. No significant
eQTL were evident at this locus.
3.1.4. MAPT
Screening of MAPT revealed four coding and five noncoding variants. All four coding variants resulted in synonymous changes: rs1052553 (Ala227Ala), rs17652121
(Asn255Asn), rs11568305 (Pro270Pro) and the novel
Gln351Gln (G>A on chromosome 17; position: 46018673,
GRCh38). We identified the known silent changes in both
cases and controls with similar MAF within our cohort,
whereas the novel Gln351Gln was identified exclusively in
one MCI patient (GM5009; MAF: 0.008). The known variants (rs1052553, rs17652121 and rs11568305) in our cohort
had similar MAF when compared to the general population.
The non-coding variant rs3744460 was only found in one
PSP patient (AC3109; MAF: 0.008) in our cohort, whereas
its MAF in the general population is 0.105 (1000G). All
other non-coding changes (rs17650901, rs62063850,
rs75534191 and rs66499584) were evenly identified in cases
and controls, and their MAF resembled that of the general
population. The eQTL analysis revealed significant effects
on transcript levels for rs17650901, rs62063850, rs1052553,
rs17652121 and rs75534191 in ten brain areas (cerebellum,
frontal cortex, hippocampus, medulla, occipital cortex, putamen, substantia nigra, temporal cortex, thalamus and white
matter). However, at the current stage, not much can be derived from this data as the haplotype analysis (see Supplementary Fig. 2) and the association analyses between the
H1/H1 and H1/H2 genotypes or the H1 and H2 alleles (see
section 3.2) did not differ when comparing AD or AD + MCI
vs. NCs. This may, in fact, suggest that no substantial differences are to be expected for transcript levels at this locus
comparatively to cases and controls in our study population.
3.1.5. GRN
Screening of GRN revealed five coding and seven noncoding variants. Two of five coding variants resulted in synonymous changes: rs25646 (Asp128Asp) that we found in 6

Current Alzheimer Research, 2015, Vol. 12, No. 8

805

cases (AC3609 [AD], LV6409 [MCI], GB6709 [MCI +
VaD], MM7309 [AD], GB8009 [AD] and FB6909 [MCI];
MAF: 0.049) and no controls, and rs140298583 (Thr409Thr)
which we identified in one control (3_P; MAF: 0.018).
Rs25646 was previously reported in control individuals [48]
and its MAF in the general population is currently reported
as 0.080 (1000G) or between 0.020 and 0.042 (dbSNP).
Three coding variants resulted in missense changes: two
(rs63750541 [Ala324Thr] and rs148213321 [Ser398Leu])
were predicted to be benign (100% of accordance) and were
identified in controls only (MDC7809 and GC2309, respectively; MAF: 0.018), although these variant is almost absent
from the general population (MAF: 0.001). MDC7809, age
at visit 71 and current age 76, and GC2309, age at visit 77
and current age 82 are currently both confirmed as healthy
controls, although GC2309 has signs of hypertension and
hypercholesterolemia. Both controls (MDC7809 and
GC2309) will need to be followed longitudinally to confirm
their disease-free status. Conversely, we found the change
rs63750412 (Arg433Trp) in one FTD case (MTP7009;
MAF: 0.008) and none of our controls. MTP7009 is a sporadic FTD case with age of onset 61. This variant was predicted to be possibly damaging (60% of accordance). Previous reports on this variant have been equivocal as it was
found in 5 FTLD cases but no controls in one study [48], one
bv FTD case and no controls in another report [49], whereas
it was found in 2/646 controls in another study [50]. This
variant has been reported in the general population with
MAF of 0.003-0.004 (1000G and dbSNP). Two of seven
non-coding changes were novel (A>G and G>T on chromosome 17; positions: 44349351 and 44349374, respectively,
GRCh38) and were only identified in two AD cases (LS8209
and AP2709, respectively; MAF: 0.008). The remainder noncoding variants (rs60100877, rs9897526, rs850713,
rs72824736 and rs5848) were evenly identified in cases and
controls, and their MAF were similar to that of the general
population. No significant eQTL were evident at this locus.
3.1.6. TARDBP
Sequencing of TARDBP revealed one coding and five
non-coding variants. The coding variant resulted in the missense change rs80356723 (Gly295Ser); we isolated this variant in one AD case (FML3209; MAF: 0.008) and no controls. FML3209 is an AD patient with age of onset 79 and
current age of 89 with very slow degenerative process and no
family history of dementia. Furthermore, this variant is absent from the general population (no MAF from 1000G and
dbSNP). Although it was predicted to be benign by our in
silico analysis (100% of accordance), it has been previously
reported in familial and sporadic French FTLD-MND (fronto
temporal lobar degeneration overlapping with motor neuron
disease) cases and absent from 400 normal controls [51], in
two sporadic Italian ALS cases (and free from >650 normal
controls) [52, 53], and now in 1 Italian AD case. Considering
that TDP-43 pathology seems a common feature in ALS,
FTD and AD, although rarer in AD [54, 55], our current and
previous findings might suggest a potential pathogenic role
for this missense change in a spectrum of neurological disorders that include ALS and AD in the Italian population. Two
of five non-coding variants (rs968545 and rs4133584) were
almost evenly identified in our cases and controls and resembled the MAF of the general population. The remainder


806 Current Alzheimer Research, 2015, Vol. 12, No. 8

variants (rs11121679, rs11121680 and rs11121681) were all
identified in one AD case (AM4209; MAF: 0.008) and none
of our controls; these three variants were however reported in
the general population with a MAF of 0.118, 0.119 and 0.12,
respectively (1000G). The eQTL analysis revealed significant
effects on transcript levels for rs4133584 in occipital cortex.
However, not much can currently be derived from this data as
the MAF and haplotype data (see further below) were not significantly different when comparing AD or AD + MCI vs.
NCs, lending support to the idea that no substantial differences
are to be expected between the transcript levels at this locus in
our cases and controls.
3.2. Haplotype Analyses and Association with Disease
Status (AD and/or AD + MCI)
No haplotype block or pair wise LD could be established
at the PSEN1, GRN and TARDBP loci (data not shown).
Only at the TARDBP locus a strong pairwise LD was detectable within the case cohort (Supplementary Fig. 3). The latter was driven by rs11121679, rs11121680 and rs11121681
that were exclusively identified in 1 AD case (AM4209),
thus such pattern reflects our sequencing results but is most
probably not informative since, as shown above, these three
variants are common polymorphisms. The structure of the
haplotype encompassing the PSEN2 locus (~12 kb) was
similar between cases and controls regardless whether AD
only or AD + MCI groups were compared to NCs (Fig. 1).
However, we identified three sub-haplotypes (A [CTACCGTA], B [CCGCCGCG] and C [TCGTCGCG]); subhaplotype B showed different frequencies comparing cases
to controls (Table 1a). None of the sub-haplotypes (A, B or
C) were significantly associated with disease status (Table
1b), however it could be inferred from our data that the subhaplotype B might be protective in our study population and

Ferrari et al.

it should be investigated in larger sample size from this geographical area.
The frequencies of the MAPT haplotype markers did not
significantly differ between cases and controls (Supplementary Table 3a). Neither the haplotypes (H1/H1 or H1/H2) nor
the alleles (H1 or H2) significantly associated with the AD
or AD + MCI groups (Supplementary Table 3b and c); this
was entirely reflected by the haplotype plot (Supplementary
Fig. 2). The haplotype blocks or pairwise LD established at
the ACE locus were equivalent in our cases and controls
(Fig. 2), and did not significantly discriminate health from
disease in our cohort.
3.3. Genotyping Results and Association with Disease
Status (AD and/or AD + MCI)
The APOE genotyping analysis showed that the 3/3
genotype was over-represented in the controls (Table 2a).
The other homozygous genotypes (2/2 and 4/4) were not
present in our study population, whereas the genotypes 2/3,
2/4 and 3/4 were always over-represented in the cases (either
AD or AD + MCI groups) when compared to controls (Table
2a); the single allele APOE*2 was more frequent in AD
(0.04) or AD + MCI (0.033) when compared to the NCs
(0.018) as well as APOE*4 (AD [0.23] or AD + MCI [0.233]
vs. NC [0.196]), whereas APOE*3 was more frequent in the
NC (0.786) compared to AD (0.730) or AD + MCI (0.733)
(Table 2a). A similar pattern had previously been reported in
the Italian population [56]. However, in our current study,
although the APOE*4 allele was more frequent in cases
(17/37 [45.9%] in the AD group, 21/45 [46.6%] in the AD +
MCI group and 11/28 [39.3%] in the NCs group), both the
association analyses, either stratified by genotype or allele,
were eventually non-significant (Table 2b and c).

Fig. (1). Polymorphisms and linkage disequilibrium at the PSEN2 locus. In each box the r2 value between 2 SNPs is shown with ranges
that vary between 0 and 0.99. Any value r2>0.8 is suggestive of LD. Comparison between the normal controls (NCs) vs. AD and AD + MCI
groups.


Screening of EOAD and LOAD Risk Factors

Table 1.

Current Alzheimer Research, 2015, Vol. 12, No. 8

807

Analysis of the PSEN2 haplotype.

a
Substructure A

Substructure B

Substructure C

CTACCGTA

CCGCCGCG

TCGTCGCG

Frequency
AD+MCI (n=45)

0.644

0.1

0.166

AD (n=37)

0.662

0.095

0.162

NC (n=28)

0.554

0.232

0.125

b

Pearson's
Chi2

Substructure A

Substructure B

Substructure C

CTACCGTA

CCGCCGCG

TCGTCGCG

Fisher's
Exact Test

OR

CI (95%)

Pearson's
Chi2

2-Tailed

Fisher's
Exact Test

OR

CI (95%)

Pearson's
Chi2

2-Tailed

Fisher's
Exact Test

OR

CI (95%)

2-Tailed

AD+MCI
(n=45) vs. NC
(n=28)

0.532

0.623

1.36

(0.5175 3.571)

/

0.168

0.36

(0.0912 1.4037)

/

0.732

1.54

(0.3624 6.5028)

AD (n=37) vs.
NC (n=28)

0.527

0.61

1.39

(0.5054 3.7935)

/

0.306

0.44

(0.1123 1.7585)

/

0.721

1.61

(0.3662 7.1045)

The three different haplotype substructures (A, B and C) at the PSEN2 locus are shown with their frequencies across the different groups (AD+MCI, AD only and NC). Substructures
A and C did not substantially differ between cases and controls; substructure C showed higher frequency in the controls comparatively to the cases groups (a). None of the subhaplotypes did significantly associate with disease status; significance threshold: p=0.05/3=0.016 after Bonferroni correction (b). Abbreviations: AD= Alzheimer’s disease; MCI=
mild cognitive impairment; NC= normal control.

Fig. (2). Polymorphisms and linkage disequilibrium at the ACE locus. In each box the r2 value between 2 SNPs is shown with ranges that
vary between 0 and 0.99. Any value r2>0.8 is suggestive of LD. Comparison between the normal controls (NCs) vs. AD and AD + MCI
groups.

Other than the case of the CR1(rs3818361), the MAF of
the risk alleles for CLU (rs11136000) and PICALM
(rs3851179) did not substantially vary when comparing
cases vs. controls (Table 3a); in addition, none of the risk
alleles significantly associated with disease status in both

comparisons (AD + MCI vs. NCs and AD only vs. NCs)
(Table 3b). Only the rs3818361 risk allele for CR1associated
with disease status (Pearson’s Chi2 test: p-value= 0.040;
OR= 2.696; Table 3b) when comparing the AD + MCI group
vs. the NCs.


808 Current Alzheimer Research, 2015, Vol. 12, No. 8

Table 2.

Ferrari et al.

Analysis of APOE Genotypes and Alleles.

a
APOE

Diagnosis

Cases

Controls

2/2

2/3

2/4

3/3

3/4

4/e4

APOE*2

APOE *3

APOE *4

AD (n=37)

/

2/37
(5.4%)

1/37
(2.7%)

18/37
(48.6%)

16/37
(43.2%)

/

0.04

0.73

0.23

MCI (n=8)

/

/

/

4/8

4/8

/

0

0.75

0.25

MCI & AD (n=45)

/

2/45
(4.4%)

1/45
(2.2%)

22/45
(48.9%)

20/45
(44.4%)

/

0.033

0.733

0.233

AD/CVD & MCI/CVD
(n=6)

/

1/6

/

3/6

2/6

/

0.083

0.75

0.167

MCI/CVD & MCI &
AD/CVD & AD (n=51)

/

3/51

1/51

25/51

22/51

/

0.039

0.735

0.225

FTD (n=8)

/

2/8

/

3/8

3/8

/

0.125

0.687

0.187

PSP (n=2)

/

/

/

2/2

/

/

0

1

0

NC (n=28)

/

1/28
(3.6%)

/

16/28
(57.1%)

11/28
(39.3)

/

0.018

0.786

0.196

b
2/3
Pearson's
Chi2 Test

3/3

Fisher's
Exact
Test

OR

CI (95%)

Pearson's
Chi2 Test

2-Tailed

3/4

Fisher's
Exact
Test

OR

CI (95%)

Pearson's
Chi2 Test

2-Tailed

Fisher's
Exact
Test

OR

CI (95%)

2-Tailed

AD+MCI (n=45)
vs. NC (n=28)

/

1

1.256

(0.1086 14.5276)

0.493

0.631

0.717

(0.2775 1.8544)

0.663

0.808

1.236

(0.4736 3.2279)

AD (n=37) vs. NC
(n=28)

/

1

1.543

(0.1328 17.9228)

0.498

0.617

0.711

(0.2646 1.9081)

0.752

0.803

1.178

(0.4336 3.1979)

c
APOE*2
Pearson's
Chi2 Test

Fisher's
Exact
Test

APOE*3

OR

CI (95%)

Pearson's
Chi2 Test

2-Tailed

Fisher's
Exact
Test

APOE*4

OR

CI (95%)

Pearson's
Chi2 Test

2-Tailed

Fisher's
Exact
Test

OR

CI (95%)

2-Tailed

AD+MCI (n=45)
vs. NC (n=28)

/

1

1.897

(0.1924 18.6958)

0.475

0.556

0.75

(0.34 1.6542)

0.603

0.683

1.245

(0.5481 2.8283)

AD (n=37) vs. NC
(n=28)

/

0.634

2.324

(0.2352 22.9588)

0.462

0.54

0.736

(0.3247 1.6701)

0.647

0.674

1.22

(0.5198 2.8638)

The frequencies and distribution of the 2/2, 2/3, 2/4, 3/3, 3/4 and 4/4 genotypes and the APOE*2, APOE*3 and APOE*4 alleles are shown for the different groups (AD+MCI, AD
only and NC) (a). Neither the E2/3, E3/3 and E3/4 genotypes (b) nor the E2, E3 and E4 alleles (c) did significantly associate with disease status; significance threshold:
p=0.05/6=0.008 after Bonferroni correction. Of note, the 2/2, 2/4 and 4/4 genotypes were not assessed as 2/2 and 4/4 were not represented in our entire study population and 2/4 was
not present in the controls. Abbreviations: AD= Alzheimer’s disease; MCI= mild cognitive impairment; NC= normal control.

4. DISCUSSION
This pilot study reports the findings of our genetic
screening of EOAD (PSEN1, PSEN2 and APP) and LOAD
(APOE, CLU, CR1 and PICALM, as well as the MAPT and
ACE haplotypes) genetic markers in an Italian cohort from a
defined geographical area (north Italy, metropolitan area of

Genoa). We also sequenced the MAPT, GRN and the
TARDBP genes in the entire cohort. The patients under study
were primarily diagnosed with AD (n = 37) and MCI (n = 8);
in addition, there were a minority of cases with concomitant
cerebrovascular disease AD + CVD (n = 3) and MCI + CVD
(n = 3) or diagnosed with FTD (n = 8) and PSP (n = 2).


Screening of EOAD and LOAD Risk Factors

Table 3.

Current Alzheimer Research, 2015, Vol. 12, No. 8

809

Analysis of CLU, CR1 and PICALM.

a
CLU

CR1

PICALM

rs11136000

rs3818361

rs3851179

Risk Allele (C)

Risk Allele (A)

Risk Allele (C)

Diagnosis

MAF
AD (n=37)

0.4

0.24

0.38

MCI & AD (n=45)

0.37

0.25

0.36

NC (n=28)

0.47

0.1

0.3

Cases

Controls

b
CLU

CR1

PICALM

rs11136000

rs3818361

rs3851179

Risk Allele (C)

Risk Allele (A)

Risk Allele (C)

Pearson's Fisher's
Chi2 Test Exact Test

OR

CI (95%)

Pearson's
Chi2 Test

2-Tailed

Fisher's
Exact
Test

OR

CI
(95%)

Pearson's
Chi2 Test

2-Tailed

Fisher's
Exact Test

OR

CI (95%)

2-Tailed

AD+MCI (n=45)
vs. NC (n=28)

0.301

0.387

1.4275

(0.7258 2.8072)

0.04

0.052

2.696

(1.0182 7.1389)

0.435

0.477

0.7529

(0.3691 1.5359)

AD (n=37) vs. NC
(n=28)

0.502

0.592

1.2711

(0.6307 2.5618)

0.07

0.103

2.485

(0.9096 6.7913)

0.374

0.457

0.716

(0.3422 1.4985)

The frequenciy of the minoralleles (MAF) of the three AD-GWAS associated SNPs are shown in the different groups (AD+MCI, AD only and NC) (a). None of the risk alleles did
significantly associate with disease status after Bonferroni correction (b); significance threshold: p=0.05/3=0.016 after Bonferroni correction. However, it is of note that the risk allele
of rs3818361 (CR1) was nominally significant in the AD+MCI group. Abbreviations: AD= Alzheimer’s disease; MCI= mild cognitive impairment; NC= normal control; GWAS=
genome wide association study.

Also, age and gender matched controls were recruited for the
purposes of this study from the same metropolitan area (NC
= 28). We characterized the entire study population and then
we: 1) interpreted the sequencing data for the entire cohort,
and 2) used data from pure AD and MCI cases for the analysis of LOAD risk factors in comparison with the NCs.
Sequencing of the EOAD candidate genes (APP, PSEN1
and PSEN2) revealed two missense changes in PSEN1 and
three in PSEN2. The PSEN1 mutations (Arg35Gln and
Glu318Gly) were predicted to be benign by the in silico
prediction-tools that we used and were reported in the general population (data from 1000G and dbSNP), but we identified them only in cases and in none of our controls. It may
be inferred that these variants might not be fully penetrant
or, rather, modulating risk. It is all the more noteworthy that
these mutations were found in AD, FTD and PSP cases,
suggesting that variability in PSEN1 encompasses wider
clinical spectrum than previously anticipated. The PSEN2
mutations (Arg62His, Ser130Leu and Tyr231Cys) are of
variable pathogenicity. Based on our data (presence in one
AD case and one control) and, in accordance with previous
reports [44, 45], the pathogenicity of the Arg62His variant
seems unclear. The Ser130Leu variant was found in one AD
+ CVD case and no controls, it was predicted to be possibly
pathogenic by the in silico prediction-tools, and it is almost

absent from the general population, thus it is potentially
pathogenic [46]. TheTyr231Cys, although predicted to be
probably pathogenic by the in silico prediction-tools, was
exclusively found in one control (MRP5309), therefore considering that this variant was previously reported as segregating with disease (bvFTD) [47] and that it is not found in
the general population one might consider the mutation as
being either non-pathogenic or pathogenic with reduced
penetrance. Further, sequencing of the other dementia genes
(MAPT, GRN and TARDBP) revealed missense changes in
GRN and TARDBP. Two missense mutations in GRN, i.e.
Ala324Thr and Ser398Leu, were identified in controls only
(MDC7809 and GC2309, respectively). These variants
were predicted to be benign by the in silico predictiontools, but they are absent from the general population. It
will be necessary to regularly follow-up these individuals
(MDC7809 and GC2309) to better interpret the relevance
of these variants. Another missense mutation in GRN,
Arg433Trp, was identified in one sporadic FTD case and
no controls. This variant was predicted to be possibly
pathogenic by the in silico prediction-tools; however, its
pathogenicity seems currently equivocal based on previous
reports [50] and the finding of a MAF of 0.003. Finally, we
found one missense change in TARDBP (Gly295Ser) in one
AD case and no controls. Although this variant was pre-


810 Current Alzheimer Research, 2015, Vol. 12, No. 8

dicted to be benign by the in silico prediction-tools, it has
been previously reported only in FTLD-MND and ALS
cases [51-53] and it is absent from the general population.
To the best of our knowledge this is the first report of this
mutation in AD, thus mutations in TARDBP might underlie
more than one clinical syndrome bridging different disorders such as FTD, ALS and, now, also AD. Of note, although AD is a clinically and pathologically moderately
homogeneous, it is becoming evident that its underlying
genetics is more complex and heterogeneous than previously thought. Based on our recent findings in PSEN1 and
TARDBP and previous reports [57] regarding existence of
pathogenic mutations in a wide spectrum of clinical neurological disorders including AD, FTD, ALS and PSP, it
might be warranted to always screen all known candidate
genes of Alzheimer’s and non-Alzheimer’s forms of dementias in all dementia cases.
The analysis of LOAD genetic risk factors highlighted
that none associated with disease status in our study population. It needs to be acknowledged that based on sample size
there was an intrinsic reduced power associated with the
current study (Supplementary Table 4a-e), thus our results
will require replication in much larger cohorts from this geographical area. We could, neither verify that the MAPT (e.g.
H1/H1 or the H1 allele) nor the ACE haplotype, nor any
APOE genotype or allele, nor the risk alleles of the GWAS
associated SNPs (rs11136000 [CLU], rs3818361 [CR1],
rs3851179 [PICALM]) associated with AD or AD + MCI
status after multiple corrections. However, two outcomes of
our analyses are preliminarily noteworthy: 1) a subhaplotype at the PSEN2 locus might exert protective effect
because of its higher frequency in controls and 2) the risk
allele of rs3818361 CR1 might associate with disease status
(nominal significance from the Pearson’s Chi2 test) in this
population. The lack of association with APOE*4 allele deserves more attention: on one hand this might just be due to
the current limited power of our study, thus a larger sample
size might lead to significant association with disease. On
the other, it needs to be kept in mind that APOE is a risk
factor thus it is neither necessary nor sufficient, alone, to
cause AD. This suggests that other factors, as it has already
been shown in the Italian population, including cerebrovascular condition, lifestyle and diet, the lipid metabolism (cholesterol homeostasis), level of education and cognitive reserve to name a few, might substantially influence and modify the incidence, progression and outcome of disease [5860]. In addition, there have been conflicting results about the
APOE*4 association with AD in the Italian population in
that in several cases concomitant comorbidities such as cerebrovascular condition or traumatic brain injury (TBI) or
hypercholesterolemia were observed [58, 61-63]. As well, a
number of studies reported certain variability in the prevalence and distribution of the APOE*4 allele, when assessed
in different geographical areas of Italy, showing a decrease
of the E4 allele frequency in Northern/central Italy compared
to the South or islands (Sicily and Sardinia) (for review see
[64]). Not least, some studies excluded unequivocal association between APOE*4 and AD in a cohort from central Italy
older than 80 [65], or another cohort form northern Italy and
older than 75 [66], or even between any APOE genotype

Ferrari et al.

and/or alleles with either longevity or disability in the Italian
population [67, 68].
One interesting outcome of our study is that the female
patients exceeded three times the male patients in the AD
(75.7% vs. 24.3%) and the AD + MCI (73.3% vs. 26.7%)
groups, and two times in the MCI (62.5% vs. 37.5%) group
(Supplementary Table 1b). It is now appreciated that there is
a substantial difference in the incidence of AD when stratified by genders as women are at a twofold risk of developing
late onset Alzheimer’s disease (LOAD) compared to men
[69], which is not considered solely due to survival rate [70].
In this respect, our data is of high interest as it might suggest
that the female gender is possibly one of the most relevant
risk factors in our study population.
CONCLUSION
In summary, keeping in mind the reduced power and the
use of a longitudinal cohort, we present here for the first time
a thorough screening of EOAD and LOAD genetic risk factors in a population from Liguria, Italy. Our preliminary data
support the finding that: 1) Known EOAD risk factors
(PSEN1 and PSEN2) as well as other genes (TARDBP)
might associate with more than one syndrome (AD, AD +
CVD, FTD and PSP) and might contribute to disease with
variable penetrance; 2) Known LOAD risk factors do not
associate with disease status (AD and/or AD + MCI) in our
study population suggesting that gene-gene and geneenvironment interaction might play a major role in the aetiology of AD in Italy, and; 3) We identified: i) a potentially
protective haplotype at the PSEN2 locus in our AD and/or
AD + MCI; ii) a nominal association with the GWAS-risk
allele A for rs3818361 in CR1 in our AD + MCI group, and;
iii) a threefold increase in disease (AD and/or AD + MCI)
susceptibility for the female population of our cohort compared to men. Our results will need to be further assessed
and confirmed in larger cohorts from this area.
SUPPLEMENTARY MATERIAL
Supplementary material is available on the publishers
web site along with the published article.
CONFLICT OF INTEREST
The authors confirm that this article content has no conflict of interest.
ACKNOWLEDGEMENTS
The authors would like to thank the patients and their
families for allowing and supporting this research. This work
was supported by the office of the Dean of the School of
Medicine, Department of Internal Medicine, at Texas Tech
Health Sciences Center.
REFERENCES
[1]

[2]

Sosa-Ortiz AL, Acosta-Castillo I, Prince MJ. Epidemiology of
dementias and Alzheimer's disease. Arch of Med Res 43(8): 600-8
(2012).
Hebert LE, Weuve J, Scherr PA, Evans DA. Alzheimer disease in
the United States (2010-2050) estimated using the 2010 census.
Neurology 80(19): 1778-83 (2013).


Screening of EOAD and LOAD Risk Factors
[3]
[4]

[5]
[6]

[7]

[8]

[9]
[10]

[11]

[12]

[13]

[14]

[15]
[16]

[17]

[18]
[19]

[20]
[21]
[22]

[23]

[24]

[25]

Thies W, Bleiler L, Alzheimer's A. 2013 Alzheimer's disease facts
and figures. Alzheimers Dementia 9(2): 208-45 (2013).
Norton S, Matthews FE, Barnes DE, Yaffe K, Brayne C. Potential
for primary prevention of Alzheimer's disease: an analysis of
population-based data. Lancet Neurol 13(8): 788-94 (2014).
Brookmeyer R, Johnson E, Ziegler-Graham K, Arrighi HM.
Forecasting the global burden of Alzheimer's disease. Alzheimers
Dementia 3(3): 186-91 (2007).
Ryan NS, Rossor MN. Correlating familial Alzheimer's disease
gene mutations with clinical phenotype. Biomarkers Med 4(1): 99112 (2010).
Thomas JB, Brier MR, Bateman RJ, Snyder AZ, Benzinger TL,
Xiong C, et al. Functional Connectivity in Autosomal Dominant
and Late-Onset Alzheimer Disease. JAMA Neurol 71(9): 1111-22
(2014).
Strittmatter WJ, Saunders AM, Schmechel D, Pericak-Vance M,
Enghild J, Salvesen GS, et al. Apolipoprotein E: high-avidity
binding to beta-amyloid and increased frequency of type 4 allele in
late-onset familial Alzheimer disease. Proc Natl Acad Sci USA
90(5): 1977-81 (1993).
Avramopoulos D. Genetics of Alzheimer's disease: recent
advances. Genome Med 1(3): 34 (2009).
Escott-Price V, Bellenguez C, Wang LS, Choi SH, Harold D, Jones
L, et al. Gene-wide analysis detects two new susceptibility genes
for Alzheimer's disease. PloS One 9(6): e94661 (2014).
Harold D, Abraham R, Hollingworth P, Sims R, Gerrish A,
Hamshere ML, et al. Genome-wide association study identifies
variants at CLU and PICALM associated with Alzheimer's disease.
Nat Genet 41(10): 1088-93 (2009).
Lambert JC, Heath S, Even G, Campion D, Sleegers K, Hiltunen
M, et al. Genome-wide association study identifies variants at CLU
and CR1 associated with Alzheimer's disease. Nat Genet 41(10):
1094-9 (2009).
Lambert JC, Ibrahim-Verbaas CA, Harold D, Naj AC, Sims R,
Bellenguez C, et al. Meta-analysis of 74,046 individuals identifies
11 new susceptibility loci for Alzheimer's disease. Nat Genet
45(12): 1452-8 (2013).
Ferrari R, Moreno JH, Minhajuddin AT, O'Bryant SE, Reisch JS,
Barber RC, et al. Implication of common and disease specific
variants in CLU, CR1, and PICALM. Neurobiol Aging 33(8): 1846
e7-18 (2012).
Momeni PF, Ferrari R. Genetic and Blood Biomarkers of
Alzheimer’s Disease. Open Nuc Med J 2: 12-24 (2010).
Allen M, Kachadoorian M, Quicksall Z, Zou F, Chai HS, Younkin
C, et al. Association of MAPT haplotypes with Alzheimer's disease
risk and MAPT brain gene expression levels. Alzheimers Res Ther
6(4): 39 (2014).
Gerrish A, Russo G, Richards A, Moskvina V, Ivanov D, Harold D,
et al. The role of variation at AbetaPP, PSEN1, PSEN2, and MAPT
in late onset Alzheimer's disease. J Alzheimers Dis 28(2): 377-87
(2012).
Ferrari R, Dawoodi S, Raju M, Thumma A, Hynan LS, Maasumi
SH, et al. Androgen receptor gene and sex-specific Alzheimer's
disease. Neurobiol Aging 34(8): 2077 e19-20 (2013).
Lunnon K, Mill J. Epigenetic studies in Alzheimer's disease:
current findings, caveats, and considerations for future studies. Am
J Med Genet Part B Neuropsychiatr Genet 162B(8): 789-99 (2013).
van den Hove DL, Kenis G, Rutten BP. Epigenetic dysregulation in
Alzheimer's disease: cause or consequence? Epigenomics 6(1): 911 (2014).
Petersen RC. Mild cognitive impairment as a diagnostic entity. J
Inter Med 256(3): 183-94 (2004).
Dubois B, Feldman HH, Jacova C, Dekosky ST, Barberger-Gateau
P, Cummings J, et al. Research criteria for the diagnosis of
Alzheimer's disease: revising the NINCDS-ADRDA criteria.
Lancet Neurol 6(8): 734-46 (2007).
Boyle PA, Wilson RS, Aggarwal NT, Tang Y, Bennett DA. Mild
cognitive impairment: risk of Alzheimer disease and rate of
cognitive decline. Neurology 67(3): 441-5 (2006).
Okello A, Koivunen J, Edison P, Archer HA, Turkheimer FE,
Nagren K, et al. Conversion of amyloid positive and negative MCI
to AD over 3 years: an 11C-PIB PET study. Neurology 73(10):
754-60 (2009).
Corradi L, Porro I, Schenone A, Momeni P, Ferrari R, Nobili F,
et al. A repository based on a dynamically extensible data model

Current Alzheimer Research, 2015, Vol. 12, No. 8

[26]

[27]
[28]
[29]

[30]

[31]

[32]

[33]
[34]

[35]

[36]
[37]

[38]

[39]
[40]

[41]

[42]

[43]
[44]

[45]

[46]

[47]

811

supporting multidisciplinary research in neuroscience. BMC Med
Informatics Decision Making 12: 115 (2012).
Folstein MF, Folstein SE, McHugh PR. "Mini-mental state". A
practical method for grading the cognitive state of patients for the
clinician. J Psychiatr Res 12(3): 189-98 (1975).
Katz S, Downs TD, Cash HR, Grotz RC. Progress in development
of the index of ADL. Gerontologist 10(1): 20-30 (1970).
Lawton MP, Brody EM. Assessment of older people: selfmaintaining and instrumental activities of daily living.
Gerontologist 9(3): 179-86 (1969).
Gorno-Tempini ML, Hillis AE, Weintraub S, Kertesz A, Mendez
M, Cappa SF, et al. Classification of primary progressive aphasia
and its variants. Neurology 76(11): 1006-14 (2011).
Litvan I, Agid Y, Calne D, Campbell G, Dubois B, Duvoisin RC,
et al. Clinical research criteria for the diagnosis of progressive
supranuclear palsy (Steele-Richardson-Olszewski syndrome):
report of the NINDS-SPSP international workshop. Neurology
47(1): 1-9 (1996).
Rascovsky K, Hodges JR, Knopman D, Mendez MF, Kramer JH,
Neuhaus J, et al. Sensitivity of revised diagnostic criteria for the
behavioural variant of frontotemporal dementia. Brain 134(Pt 9):
2456-77 (2011).
Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and
visualization of LD and haplotype maps. Bioinformatics 21(2):
263-5 (2005).
Adzhubei IA, Schmidt S, Peshkin L, Ramensky VE, Gerasimova
A, Bork P, et al. A method and server for predicting damaging
missense mutations. Nat Meth 7(4): 248-9 (2010).
Kumar P, Henikoff S, Ng PC. Predicting the effects of coding nonsynonymous variants on protein function using the SIFT algorithm.
Nat Protocol 4(7): 1073-81 (2009).
Thomas
PD,
Kejariwal
A.
Coding
single-nucleotide
polymorphisms associated with complex vs. Mendelian disease:
evolutionary evidence for differences in molecular effects. Proc
Natl Acad Sci USA 101(43): 15398-403 (2004).
Yue P, Melamud E, Moult J. SNPs3D: candidate gene and SNP
selection for association studies. BMC Bioinform 7: 166 (2006).
Frederic MY, Lalande M, Boileau C, Hamroun D, Claustres M,
Beroud C, et al. UMD-predictor, a new prediction tool for
nucleotide substitution pathogenicity -- application to four genes:
FBN1, FBN2, TGFBR1, and TGFBR2. Hum Mut 30(6): 952-9
(2009).
Raux G, Guyant-Marechal L, Martin C, Bou J, Penet C, Brice A,
et al. Molecular diagnosis of autosomal dominant early onset
Alzheimer's disease: an update. J Med Genet 42(10): 793-5 (2005).
Rogaeva EA, Fafel KC, Song YQ, Medeiros H, Sato C, Liang Y,
et al. Screening for PS1 mutations in a referral-based series of AD
cases: 21 novel mutations. Neurology 57(4): 621-5 (2001).
Aldudo J, Bullido MJ, Frank A, Valdivieso F. Missense mutation
E318G of the presenilin-1 gene appears to be a nonpathogenic
polymorphism. Ann Neurol 44(6): 985-6 (1998).
Dermaut B, Cruts M, Slooter AJ, Van Gestel S, De Jonghe C,
Vanderstichele H, et al. The Glu318Gly substitution in presenilin 1
is not causally related to Alzheimer disease. Am J Human Genet
64(1): 290-2 (1999).
Taddei K, Fisher C, Laws SM, Martins G, Paton A, Clarnette RM,
et al. Association between presenilin-1 Glu318Gly mutation and
familial Alzheimer's disease in the Australian population. Mol
Psychiatry 7(7): 776-81 (2002).
Goldman JS, Johnson JK, McElligott K, Suchowersky O, Miller
BL, Van Deerlin VM. Presenilin 1 Glu318Gly polymorphism:
interpret with caution. Arch Neurol 62(10): 1624-7 (2005).
Ertekin-Taner N, Younkin LH, Yager DM, Parfitt F, Baker MC,
Asthana S, et al. Plasma amyloid beta protein is elevated in lateonset Alzheimer disease families. Neurology 70(8): 596-606
(2008).
Sleegers K, Roks G, Theuns J, Aulchenko YS, Rademakers R,
Cruts M, et al. Familial clustering and genetic risk for dementia in
a genetically isolated Dutch population. Brain 127(Pt 7): 1641-9
(2004).
Tedde A, Nacmias B, Ciantelli M, Forleo P, Cellini E, Bagnoli S, et
al. Identification of new presenilin gene mutations in early-onset
familial Alzheimer disease. Arch Neurol 60(11): 1541-4 (2003).
Marcon G, Di Fede G, Giaccone G, Rossi G, Giovagnoli AR,
Maccagnano E, et al. A novel Italian presenilin 2 gene mutation


812 Current Alzheimer Research, 2015, Vol. 12, No. 8

[48]

[49]

[50]

[51]

[52]

[53]

[54]

[55]

[56]

[57]
[58]

Ferrari et al.

with prevalent behavioral phenotype. J Alzheimers Dis 16(3): 50911 (2009).
Gass J, Cannon A, Mackenzie IR, Boeve B, Baker M, Adamson J,
et al. Mutations in progranulin are a major cause of ubiquitinpositive frontotemporal lobar degeneration. Human Mol Genet
15(20): 2988-3001 (2006).
Spina S, Murrell JR, Huey ED, Wassermann EM, Pietrini P,
Baraibar MA, et al. Clinicopathologic features of frontotemporal
dementia with progranulin sequence variation. Neurology 68(11):
820-7 (2007).
van der Zee J, Le Ber I, Maurer-Stroh S, Engelborghs S, Gijselinck
I, Camuzat A, et al. Mutations other than null mutations producing
a pathogenic loss of progranulin in frontotemporal dementia.
Human Mut 28(4): 416 (2007).
Benajiba L, Le Ber I, Camuzat A, Lacoste M, Thomas-Anterion C,
Couratier P, et al. TARDBP mutations in motoneuron disease with
frontotemporal lobar degeneration. Ann Neurol 65(4): 470-3
(2009).
Corrado L, Ratti A, Gellera C, Buratti E, Castellotti B, Carlomagno
Y, et al. High frequency of TARDBP gene mutations in Italian
patients with amyotrophic lateral sclerosis. Human Mut 30(4), 68894 (2009).
Del Bo R, Ghezzi S, Corti S, Pandolfo M, Ranieri M, Santoro D, et
al. TARDBP (TDP-43) sequence analysis in patients with familial
and sporadic ALS: identification of two novel mutations. Eur J
Neurol 16(6): 727-32 (2009).
Josephs KA, Whitwell JL, Weigand SD, Murray ME, Tosakulwong
N, Liesinger AM, et al. TDP-43 is a key player in the clinical
features associated with Alzheimer's disease. Acta Neuropathol
127(6): 811-24 (2014).
Neumann M, Sampathu DM, Kwong LK, Truax AC, Micsenyi
MC, Chou TT, et al. Ubiquitinated TDP-43 in frontotemporal lobar
degeneration and amyotrophic lateral sclerosis. Science 314(5796):
130-3 (2006).
Sorbi S, Nacmias B, Forleo P, Latorraca S, Gobbini I, Bracco L,
et al. ApoE allele frequencies in Italian sporadic and familial
Alzheimer's disease. Neurosci Lett 177(1-2): 100-2 (1994).
Galimberti
D,
Scarpini
E.
Behavioral
Genetics
of
Neurodegenerative Disorders. Curr Top Behav Neurosci 12: 61531 (2012).
Borroni B, Grassi M, Costanzi C, Archetti S, Caimi L, Padovani A.
APOE genotype and cholesterol levels in lewy body dementia and
Alzheimer disease: investigating genotype-phenotype effect on
disease risk. Am J Geriatr Psychiatry 14(12): 1022-31 (2006).

Received: February 04, 2015

[59]

[60]

[61]

[62]

[63]
[64]

[65]

[66]

[67]

[68]

[69]
[70]

Revised: May 27, 2015

Corbo RM, Scacchi R. Apolipoprotein E (APOE) allele distribution
in the world. Is APOE*4 a 'thrifty' allele? Ann Human Genet 63(Pt
4): 301-10 (1999).
Panza F, Seripa D, D'Onofrio G, Frisardi V, Solfrizzi V, Mecocci
P, et al. Neuropsychiatric Symptoms, Endophenotypes, and
Syndromes in Late-Onset Alzheimer's Disease: Focus on APOE
Gene. Intern J Alzheimers Dis 2011: 721457 (2011).
Folin M, Baiguera S, Conconi MT, Di Liddo R, De Carlo E,
Parnigotto PP, et al. Apolipoprotein E as vascular risk factor in
neurodegenerative dementia. Intern J Mol Med 14(4): 609-13
(2004).
Mauri M, Sinforiani E, Bono G, Cittadella R, Quattrone A, Boller
F, et al. Interaction between Apolipoprotein epsilon 4 and
traumatic brain injury in patients with Alzheimer's disease and
Mild Cognitive Impairment. Funct Neurol 21(4): 223-8 (2006).
Panza F, D'Introno A, Colacicco AM, Basile AM, Capurso C,
Kehoe PG, et al. Vascular risk and genetics of sporadic late-onset
Alzheimer's disease. J Neural Transm 111(1): 69-89 (2004).
Bosco P, Gueant-Rodriguez RM, Anello G, Spada RS, Romano A,
Caraci F, et al. Allele epsilon 4 of APOE is a stronger predictor of
Alzheimer risk in Sicily than in continental South Italy. Neurosci
Lett 388(3): 168-72 (2005).
Scacchi R, De Bernardini L, Mantuano E, Donini LM, Vilardo T,
Corbo RM. Apolipoprotein E (APOE) allele frequencies in lateonset sporadic Alzheimer's disease (AD), mixed dementia and
vascular dementia: lack of association of epsilon 4 allele with AD
in Italian octogenarian patients. Neurosci Lett 201(3): 231-4
(1995).
Benedetti MD, Salviati A, Filipponi S, Manfredi M, De Togni L,
Gomez Lira M, et al. Prevalence of dementia and apolipoprotein e
genotype distribution in the elderly of buttapietra, verona province,
Italy. Neuroepidemiology 21(2): 74-80 (2002).
Bader G, Zuliani G, Kostner GM, Fellin R. Apolipoprotein E
polymorphism is not associated with longevity or disability in a
sample of Italian octo- and nonagenarians. Gerontology 44(5): 2939 (1998).
Panza F, Solfrizzi V, Colacicco AM, Basile AM, D'Introno A,
Capurso C, et al. Apolipoprotein E (APOE) polymorphism
influences serum APOE levels in Alzheimer's disease patients and
centenarians. Neuroreport 14(4): 605-8 (2003).
Association As. 2012 Alzheimer’s Disease Facts and Figures.
2012.
Vina J, Lloret A. Why women have more Alzheimer's disease than
men: gender and mitochondrial toxicity of amyloid-beta peptide. J
Alzheimers Dis 20(2): S527-33 (2010).

Accepted: June 17, 2015



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

Tải bản đầy đủ ngay

×