To cite this article: Neuroendocrinol Lett 2014;35(5):359–366
REVIEWARTICLE
Neuroendocrinology Letters Volume 35 No. 5 2014Neuroendocrinology Letters
ISSN: 0172-780X; ISSN-L: 0172-780X; Electronic/Online ISSN: 2354-4716
Web of Knowledge / Web of Science: Neuroendocrinol Lett
Pub Med / Medline: Neuro Endocrinol Lett
Perspectives in genetic prediction
of Alzheimer’s disease
Omar Šerý1,2, Jana Povová3, Vladimir J. Balcar4
1 Laboratory of neurobiology and molecular psychiatry, Department of Biochemistry, Faculty of
Science, Masaryk University, Brno, Czech Republic
2 Laboratory of Animal Embryology, Institute of Animal Physiology and Genetics, the Academy of
Sciences of the Czech Republic, Brno, Czech Republic
3 Department of Epidemiology and Public Health, Faculty of Medicine, University of Ostrava, Czech
Republic
4 School of Medical Sciences (Anatomy and Histology) and Bosch Institute, Sydney Medical School,
The University of Sydney, Australia
Correspondence to: Assoc. Prof. RNDr. Omar Šerý, PhD.
Laboratory of Neurobiology and Molecular Psychiatry,
Department of Biochemistry, Faculty of Science, Masaryk University,
Kotlářská 2, 611 37 Brno, Czech Republic.
tel: +420 549 497 312; fax: +420 542 213 827; e-mail: omarsery@sci.muni.cz
Submitted: 2014-07-04 Accepted: 2014-07-14 Published online: 2014-09-28
Key words: predictive testing; gene; polymorphism; multifactorial disease; neural network
Neuroendocrinol Lett 2014;35(5):359–366 PMID: 25275266 NEL350514R01 ©2014 Neuroendocrinology Letters • www.nel.edu
Abstract Alzheimer disease (AD) represents a group of multifactorial disorders characterized
by a progressive decline of mental faculties eventually leading to dementia
and death. Aging of human populations is behind the rapid worldwide increase
in the prevalence of AD in recent decades. AD prevention critically depends on
reliable AD-predictive genetic testing but its further development is delicately
poised at present.
New DNA-analyzing technologies such as the Next Generation Sequencing (NGS)
have allowed rapid and comprehensive analysis of the genome and might have
aided the research into the genetics of AD. However, discoveries of epigenetic
mechanisms and non-coding forms of DNA and RNA – while helping to explain
complexities of AD etiologies – have imposed additional challenges onto the
AD diagnostics based on DNA analyses. Environmental factors can, via epigenetic
mechanisms, modify both coding and non-coding DNA and this has to be
respected in DNA testing, including NGS.
Risk calculations based on the known odds and risk ratios for selected DNA
polymorphisms are viable options at present, while the applications of neural
network methodology seems the most promising way forward in the development
of predictive AD tests in future.
Abbreviations:
ACE - angiotensin-converting enzyme
AD - Alzheimer`s disease
ADRA2A - α2-adrenergic receptor
APBB2 - β-amyloid precursor protein-binding family B
member 2
ApoE - apolipoprotein E
APP - amyloid precursor protein
BRCA1 - breast cancer 1 gene
CYP2D6 - cytochromeP450,family2,subfamilyD,polypeptide6
DAPK1 - death associated protein kinase 1
IDE - insulin-degrading enzyme
DNA - deoxyribonucleic acid
DRD3 - dopamine D3 receptor
GWAS - genome-wide association study
KNS2 - kinesin light chain 1 (KLC1)
mRNA - messenger ribonucleic acid
MTHFR - methylenetetrahydrofolate reductase
ncRNA - non coding ribonucleic acid
360 Copyright © 2014 Neuroendocrinology Letters ISSN 0172–780X • www.nel.edu
Omar Šerý, Jana Povová, Vladimir J. Balcar
NGS - next generation sequencing
OR - odds ratio
RNA - ribonucleic acid
RR - risk ratio
SNAP-25 - synaptosomal-associated protein of 25kDa
SNP - single-nucleotide polymorphism
TOMM40 - translocase of the mitochondrial outer membrane 40
INTRODUCTION
Alzheimer’s disease is caused by both genetic and
environmental factors. It is estimated that 35.6 million
people lived with dementia worldwide in 2010, with
numbers expected to almost double every 20 years; to
65.7 million in 2030 and 115.4 million in 2050 (Prince
et al. 2013). Most people with dementia live in developing
countries (Ferri et al. 2005).
Alzheimer’s disease can have a profound impact on
the human quality of life in affected individuals and
their families with a potential to disrupt fine social and
economic fabric of whole communities in a very near
future. Effective approach to solving this urgent problem
would require much greater level of knowledge of
both risk factors and etiologies of Alzheimer’s disease
than what is available to health professionals at present
(Povová et al. 2012). Accordingly, current research in
the area focuses mainly on the roles of genes and environment
as well as life style factors in the pathogenesis
of Alzheimer’s disease (Šerý et al. 2013). In recent years,
a possibility emerged of genetic prediction of certain
types of Alzheimer’s disease; if this approach could be
expanded over as many potential patients as possible,
patients at high risk of the disease could be identified
at the level of general practice and preventive measures
could be instituted well before the onset of the actual
symptoms.
The aim of this paper is not to provide a comprehensive
review of genetic or non-genetic mechanisms
of AD pathology. Nor do we intend to present every
fine engineering detail of the DNA analyzing technologies
that have recently appeared on the scene (several
specific methods are critically discussed but mainly in
the light of their potential impact on predictive diagnostics
of AD). Rather, we wish to briefly overview the
current knowledge relevant to the assessment of the
risks of Alzheimer’s disease particularly with respect
to the current trends in DNA analysis and its use in
diagnostics.
In the first section we try to capture the most recent
developments in DNA-analyzing techniques and provide
some critical comments as to their potential use
in genetic testing for the risk of multifactorial diseases.
The second section discusses the hurdles faced by
genetic testing in psychiatry; this includes subsections
on epigenetics and non-coding forms of DNA and RNA
and their possible relevance in the etiology of complex
mental diseases. In the final two sections we use examples
selected from recent literature and unpublished
data to illustrate the present state of the art in identifying
the risks of multifactorial diseases. At this point we
look into the future and try to indentify the approach
most likely to reach the ultimate goal: reliable estimation
of AD risks years before the onset.
THE DEVELOPMENT OF NEW DNA
TECHNOLOGIES
The  Human Genome Project  commenced in 1990
as a collaboration of international researchers whose
shared goal was to map and understand complete
human genome. This was an extremely challenging
task at the time considering that the human genome
contains thousands of genes and 3.2 billion nucleotide
bases. The project was originally designed to take fifteen
years; however, thanks to rapid developments in
DNA technology, it was completed much earlier. The
first “draft” of the human genome was published in
February 2001 and in April 2003 sequencing of the full
human genome was finalized and published. This was
a nominally very successful outcome of the project.
However, the initial euphoria was quickly replaced by
a sobering realization that the knowledge of the whole
human genome was leaving us at considerable distance
from answering some of the most obvious and important
questions. This included even those questions
concerning apparently straightforward diagnostics of
genetically based disorders. In human DNA sequences,
thousands of new genes were discovered whose role was
– and in many cases still remains – totally obscure. Furthermore,
it was realized that, in order to understand
whether and/or how various DNA polymorphic sites
contribute to the mechanisms of genetically based diseases,
it would be necessary to read the DNA of many
thousands of individuals. Capillary sequencing technology,
used in the human genome project was clearly
not up to the task and would not, purely in terms of
experimental time, if not labour, allow realization of
such projects.
In 2002, Illumina Company introduced a new and
improved device that was capable of analysing 1536
single nucleotide polymorphisms (SNPs) simultaneously
from 96 individual human DNA specimens (Oliphant
et al. 2002). The most distinguishing feature of
the technology was a possibility of individual design
options while searching for potentially relevant polymorphisms
in each individual DNA specimen. Illumina
uses BeadChip platform that is based on a microelectro-mechanical
system in which wells are created
through a combination of photo lithography and
plasma etching on silicon wafers (Steemers & Gunderson
2007). The main concept of array manufacturing is
that DNA-immobilized beads are randomly dispersed
and assembled into wells on a slide. A decoding process
maps the location and identity of each bead on the
array. Currently, BeadArray technology uses 3-micron
silica beads that self-assemble in micro-wells on either
361Neuroendocrinology Letters Vol. 35 No. 5 2014 • Article available online: http://node.nel.edu
Alzheimer`s disease
of two substrates: fiber optic bundles or planar silica
slides. When randomly assembled on one of these two
substrates, the beads have a uniform spacing of ~5.7
microns (Steemers & Gunderson 2007). Each bead is
covered with hundreds of thousands of copies of a specific
oligonucleotides that act as the capture sequences
in one of Illumina’s assays. In a recent version of the
design, up to 2.5 millions of SNPs can be identified and
evaluated in DNA samples from eight persons, the process
taking altogether two days.
In 1998 a new type of DNA sequencing called “pyrosequencing”
was described by Ronaghi et al. (Ronaghi
et al. 1998). This methodology formed the basis of the
next generation sequencing (NGS) as offered by three
companies on semi-automatic instruments at present.
The system named GS Junior System (Roche Diagnostics
Corporation) is a platform suitable not only for
research but also for laboratories specializing in clinical
investigations. In 2013, Roche announced that GS
Junior System platform would be discontinued in or
after 2015. Illumina Company improved next generation
sequencing method originally designed by Solexa
Company and an Illumina’s high throughput sequencing
system is now available on MiSeq, NextSeq and
HiSeq instruments. New reagents enable the generation
of up to 1 terabase (Tb) of data in high output mode on
HiSeq instrument, supporting what is to date the greatest
number of samples per run. Entire human genome
can be sequenced by Illumina’s system in one run.
The Ion Proton™ System (Life Technologies) is the
first benchtop sequencing system capable of humanscale
genome, exome, or transcriptome sequencing in
a few hours. One human genome (up to 20× coverage)
can be read per run. The system combines semiconductor
sequencing technology with classical biochemistry
to directly translate chemical information into digital
data. Life Technologies Company also offers instrument
named PGM™ System  for semiconductor sequencing
up to 1 Gb. The difference between instruments of
different companies is not only in the technology and
cost but also in the accuracy and reliability of the DNA
reading. Some of these systems may not be particularly
suited for routine DNA diagnostics, though.
New, more rapid and cheaper technologies of DNA
sequencing that should appear in the next few years
have been announced and are eagerly expected. In fact,
the development in the field of molecular methodologies
is so rapid that the NGS technology sometimes
appears “to steal” the current and even future DNA
chip technology. We can expect that all DNA chip technologies
in human genome research would be soon
substituted by NGS instruments offering favourable
prices and more accurate and reliable results.
These new high throughput technologies have been
used in the investigation of many multifactorial diseases
including Alzheimer’s  disease. Former studies of socalled
candidate genes that were selected on the basis of
what we knew at the time about the pathogenesis of the
disease, have been giving way to the powerful modern
DNA chip technologies mentioned above. DNA chip
technologies have been, of course, behind the proliferation
of genome-wide association studies (GWAS).
GWAS, initially a very promising approach, was first
applied to Alzheimer’s disease in 2009. GWAS studies
compare genotype and allele frequencies of thousands
to millions polymorphisms (SNPs) in groups of thousands
of human subjects and, in principle, they should
be able to identify relevant genes by comparing patients
and controls. The initial promise of GWAS has not
been, however, always fulfilled and the problem grew in
complexity. Specific difficulties with the interpretation
of the results as encountered in psychiatric research
will be discussed later in this publication. It general
terms, it seems that GWAS based on mere correlation
of thousands DNA polymorphisms with corresponding
pathology data sets obtained from patients and control
subjects but lacking a specific prior concept may not
bring rapid progress as originally expected (Medway
& Morgan 2014; Hosák 2013). Furthermore, as new
results from whole genome genotyping of thousands
of Alzheimer‘s patients have been pouring in, the roles
of non-coding RNA and the contribution of epigenetic
factors in pathogenesis of Alzheimer‘s  disease have
been neglected.
DNA DIAGNOSTICS IN THE PSYCHIATRY
DNA diagnosis in psychiatry is not routinely used.
There are several reasons:
First reason is that mental disorders are multifactorial
diseases whose pathogenesis is most often caused
by interplay of genes and environment (Povová et al.
2012, Šerý et al. 2013). Very similar or even identical
genetic dispositions may thus react very differently to
near-identical environmental stimuli (vide infra in the
section on epigenetics). In actual fact, in multifactorial
diseases, even when the inherited predisposition is
definitely present, the environment and life-style factors
may override the genes and exert decisive influence
in the development of the pathogenesis (for a recent
review of the pathogenesis of a relatively “simple” mutifactorial
disease illustrating this last point see Sarantzis
& Bown 2014).
Epigenetics
Epigenetics has been a fast-moving field of scientific
research and, in recent years, epigenetic influences
on the pathogenesis of multifactorial diseases have
been studied with ever increasing frequency. What
are the most common and/or most likely underlying
mechanisms?
DNA-carried information is primarily encoded by
the sequence of the four constituent nucleotide bases;
this inherited genetic template (genotype) can, however,
be modified during the life-time by externallytriggered
chemical alterations of the existing DNA
362 Copyright © 2014 Neuroendocrinology Letters ISSN 0172–780X • www.nel.edu
Omar Šerý, Jana Povová, Vladimir J. Balcar
structure. Such changes can introduce completely new
(environment-related i.e. “epigenetic”) information into
the DNA molecule and this may be of decisive importance
for the final outcome: the individual’s phenotype
including his/her susceptibility to particular diseases.
One of the most often discussed epigenetic mechanisms
is the methylation of cytosine base (Wu & Zhang
2014). In fact, the methylated cytosine can be viewed
as a fifth species of nucleotide carrying whole new set
of acquired “instructions” on how to express (or not to
express) the existing genes; – indeed, generally speaking,
an increased cytosine methylation in a gene would
tend to lower the gene expression (Zhou 2012).
Cytosine methylation is affected by a wide range of
factors: age (age increases amount of methylated cytosine
in the DNA), gender (males have more methylated
cytosine), race (blacks have more methylated cytosine)
just to name a few of the factors (Issa 2014; MartínSubero
2011). The cytosine methylation is tissue specific,
which means that the same person may have the
same gene methylated differently in different tissues.
Cytosine methylation, however, should not be viewed
simply as a function of age, race, gender or a location
in a particular organ but also with respect to the organisms’s
total life history, i.e being determined, at least
in part, by a variety of specific external factors, acting
constantly, intermittently, or, in some cases, for short
periods of time but leading to lasting changes which
may produce significant effects months or years later.
Thus cytosine methylations are known to be affected
by food (different types of dietary saturated or unsaturated
fatty acids affect the degree of methylation of
cytosine), dietary amount of vitamins B6 and B12, dark
green vegetables in the diet (increases DNA methylation),
but also by stress and toxic substances present
in the environment (mercury, lead, and pesticides); all
affecting the DNA-(cytosine-) methylation status of
the DNA in living cells (Langley-Evans 2014; Maloney
et al. 2012; Chouliaras et al. 2010; DiLorenzo &
DiLorenzo 2013). Furthermore, the methylation of
cytosine can be, in part (18 – 23%), passed on to subsequent
generations (Bell et al. 2012; Boks et al. 2009)
thus providing a mechanism for inheriting acquired
characteristics.
There are other epigenetic mechanisms in play in
addition to cytosine methylation, for example the methylation
and acetylation of histones; these proteins form
specifically structured multimeres which complex with
DNA and significantly influence gene expression (Wu
2014). Overall impact of epigenetic mechanisms on
quality of life and health outcomes in humans might be
assessed, or, at least, conveniently illustrated by comparing
two individuals with identical primary DNA
structure (i.e. monozygotic twins) and note the differences
between them and their life histories, particularly
in terms of aging and disease.
Within psychiatric research, epigenetic effects are
best documented in anxiety disorders and affective
disorders. For example a paradigm of an unpredictable
maternal separation in early postnatal life has been
used to study the impact of stress experienced by very
young developing rats on their behaviour later in their
lives. The early maternal separation severely affected
behavior across several generations. Pups subjected to
this form of stress (F1 generation), and their offspring
(F2 generation obtained by breeding F1 animals to
naive controls), developed depressive-like behavior
patterns when adult, and showed deficits in a range of
specific tests such as those evaluating novelty response,
risk assessment and social interactions (Bohacek et al.
2013). These behavioural aberrations are accompanied
by persistent changes in the phenotype, detected at the
molecular level, in particular, deficiencies in components
of the stress pathway and serotonergic signaling.
Transmission of the changes and accompanying behavioural
patterns occurs through both females and males
and is observed down to the third generation (Bohacek
et al. 2013; Mychasiuk et al. 2013). Analogous (putatively
epigenetic) effects of maternal stress have also
been reported in women that lived through the Holocaust
in a concentration camp. The stressful insults
to which they had been subjected appear to influence
behavior of their offspring down to the third generation
(Lehrner et al. 2014).
From this perspective, it seems necessary that the
research on Alzheimer’s disease etiology takes into
account also the epigenetic mechanisms such as the
methylation of cytosine which has a definite potential
to affect the expression of amyloid precursor protein,
Tau protein and other proteins thought to be important
in the pathogenesis of Alzheimer’s disease (Maloney
et al. 2012; Chouliaras et al. 2010). NGS methods will
actually provide us with a possibility to detect the presence
of methyl groups on cytosine bases thus potentially
enabling the NGS instruments to include epigenetics of
multifactorial diseases (such as psychiatric disorders)
(Sarda & Hannenhalli 2014) in the research and diagnostics
procedures. One obvious limitation, however, is
the fact that the ideal samples for such analyses need to
be taken directly from human brain and not as blood
samples; this makes this approach somewhat complicated
particularly from the point of view of preventive
diagnostics.
Non-coding DNA and RNA
There is another reason, apart from the multifactorial
nature of mental disorders, why DNA diagnostics in
the psychiatry is not currently much in use; most of
the gene regulatory processes in the brain are thus far
unknown or not adequately understood. In addition
to the epigenetic mechanisms, the recent years have
brought totally new and revolutionary information on
the structure of the genetic material and the mechanisms
necessary for orderly gene expression. This, as
some of the major advances mentioned in previous
sections, was made possible by the applications of new
363Neuroendocrinology Letters Vol. 35 No. 5 2014 • Article available online: http://node.nel.edu
Alzheimer`s disease
technologies such as the modern methods using the
DNA microarrays. Each cell produces vast amounts of
RNA and not all of this RNA serves as a template for
the production of proteins (Moss 2002). Thus hitherto
unknown species of RNA called “non-coding RNAs”
have been discovered. The main role of these noncoding
RNAs seems to be to intervene in the regulation
of gene expression. The discovery of non-coding
RNA also explained, at least in part, the presence of
abundant “junk” DNA in the genome of highly complex
organisms including humans. New technologies
began to provide answers to the question: Why do
the complex organisms have the excess of noncoding
(“junk”, “waste”) DNA in the genome? It has become
apparent that most of the human  genome  is transcribed
into non-coding RNAs (ncRNAs) rather than
protein-coding  mRNAs (Shi et al. 2013). Multiple
types of ncRNAs are present in abundant quantities in
the brain, and this large group of molecules may be
involved in the development, maintenance and regulation
of the molecular and cellular complexity of the
brain; its multiple types of neurons and their relationship
with glial cells, the arrangements of vast numbers
of synapses and the dynamics behind the changes
in their characteristics and connectivities. Since the
modulation of synapses is thought to be the basis of
the mechanism of long term memory storage (Earls
et al. 2014) such phenomena – including the potential
role of ncRNAs – would be of great interest in the
Alzheimer’s disease research.
What impact has the discovery of non-coding RNA
on molecular biology and on molecular psychiatry in
particular? One immediately obvious consequence
is that in future it will be necessary to reinterpret the
current results of studies analyzing the relationship
between DNA polymorphisms and mental disorders
in the light of the emerging knowledge of non-coding
RNAs. Many DNA polymorphisms are located in DNA
areas that affect the function of non-coding RNA. This
means that polymorphisms occurring outside the classical
regulatory regions of genes and exons will have to
be taken into account. Polymorphisms may affect not
only the process of the formation of non-coding RNAs
themselves, but also their sequence and the sequence
of the target DNA structures. What compounds the
matter even further is that epigenetics including the
cytosine methylation affects the formation of ncRNA
as well and, in turn, ncRNA affects methylation of
cytosine (Miao et al. 2013). Thus the involvement of
ncRNAs adds an extra element of complexity to how
the gene expression may be subject to external influences
from the environment.
More general problem faced by DNA diagnostics
in psychiatry is the use of heterogeneous groups of
persons used in psychiatric research on humans. In
research on mice and rats, it is possible to more or
less standardize the experimental model genetically
(using the same strain of animals throughout the
experiment) and the environmental conditions such
as lighting, food, temperature, exposure to stressors
or toxic substances could be modified and controlled.
In research conducted on people such precise control
cannot be achieved. The research on a population of
persons is therefore exposed to considerable variations
in the groups of studied persons – not just genetic
(e.g. race or ethnicity), but also in terms of environmental
conditions, diet, stress, social situation, etc. all
with a potential to inadvertently introduce unknown
confounding factors. It is then not surprising that the
results of the GWAS studies may produce apparently
positive data on relationship between hundreds of
DNA polymorphisms and the pathogenesis of studied
mental disorders while other studies in other cohorts
of patients will fail to replicate them (Simundic 2010).
In our association study of schizophrenia we re-tested
in our group of patients tens of DNA polymorphic
markers described in GWAS studies but we detected
only the relationship between schizophrenia and
previously described candidate genes like ADRA2A,
DRD3, MTHFR and SNAP-25 (Lochman et al. 2013a;
Lochman et al. 2013b; Šerý et al. 2010) being unable
to replicate any associations of genes identified by the
recent GWAS.
Another crucial problem, why the diagnostics of
psychiatric disease using DNA analysis is unlikely to
produce clear-cut results is the fact that mental disorders,
as currently defined, almost certainly represent
a set of diseases with different pathogenesis, just displaying
similar symptoms. Approached from this angle
(rather than phenomenologically as is usually done in
psychiatry), it can be expected that there are dozens of
types of schizophrenia not yet described and defined in
the literature and even Alzheimer’s disease should not
be considered a single disease (Šerý et al. 2013). The
pathogenesis of Alzheimer’s disease may be the result
of multiple causes that may have escaped attention
since the recent genetic research has not adequately
focused on subtle differences in Alzheimer’s disease
pathophysiology. In Alzheimer’s disease the accumulation
of beta amyloid and phosphorylated Tau protein
may not be the primary cause of Alzheimer’s disease
(Šerý et al. 2013). Indeed, studies in recent years have
produced findings suggesting that Alzheimer’s disease
is related to stress changes in endoplasmic reticulum,
brain microhaemorrages, etc. (Šerý et al. 2013; Vianna
et al. 2012).
RECENT TRENDS IN GENETIC
PREDICTION OF MULTIFACTORIAL
DISEASES
DNA diagnostics in pharmacogenetics is currently
the only molecular diagnostics used in psychiatry
(Zhang & Malhotra 2013). Currently it is possible to
evaluate drug metabolism through genetic analysis
of CYP2D6 gene and other mitochondrial enzymes
364 Copyright © 2014 Neuroendocrinology Letters ISSN 0172–780X • www.nel.edu
Omar Šerý, Jana Povová, Vladimir J. Balcar
(Altar et al. 2013). Genetic screening can estimate
the optimum dosage of drug for each patient. For
example, fast metabolizers may need higher doses of
drugs than slow metabolizers. In the near future we
may expect the use of pharmacogenetics in another
way, specifically in predicting more accurately each
patient’s response to treatment. Genetic  studies have
been successful in identifying of DNA polymorphisms
responsible for antipsychotic-induced weight gain and,
to some extent, for the predilection to develop tardive
dyskinesia (Müller et al. 2013). In future, the analyses
of DNA polymorphisms may fall under the scope of
so-called personalized medicine and be used to predict
the metabolism, safety and efficacy of psychotherapeutic
agents thus helping to design specific treatments for
individual patients.
Recently, it has become possible to use genetic testing
in estimating predisposition of patients for certain
conditions including multifactorial diseases. This
approach is, strictly speaking, not “diagnostic” since
it merely helps to estimate a relative probability (risk)
of the patient developing a particular disease at some
future point in time. Detection of genetically determined
risk factors by analyzing DNA is probably the
only practical application of genetic testing in the near
future. The usefulness of the testing has now become
recognized to the extent that it is, in some cases, covered
by health insurance. The tests on offer include, for
example, analysis of trombophilic factors looking for
mutations associated with increased risks in thrombotic
events, abortions etc. Another example is the use
of DNA analysis in the calculation of breast cancer risk
induced by mutations in the BRCA1 gene. This has
resulted in highly publicized cases of preventive breast
ablation to reduce the risk of developing cancer and
may have already saved many lives.
Research into the molecular and cellular mechanisms
of Alzheimer disease (AD) seems somewhat
ahead of the research into other mental disorders; all
patients with Alzheimer’s disease have been identified
as having increased levels of Abeta protein in the brain.
The presence of Lewy bodies and their constituent proteins
may have been associated with idiopathic Parkinsonism
and the accompanying type of dementia (for a
review see Beitz 2014) while altered conformations of
prion proteins have been found in certain encephalopathies
(for a review see Collinge 2001). However,
the links between the presence of such structures and
the etiology of the disorders is far more tenuous than in
the case of Abeta protein and AD. Apart from AD there
are really no other commonly occurring major mental
conditions such as schizophrenia and bipolar disorder
which would have any tangible cellular or molecular
common denominator that could serve as diagnostic
tools. This somewhat simplifies the research into the
mechanisms of AD, despite the details of the pathogenesis
of AD remaining poorly understood.
CURRENT AND FUTURE DEVELOPMENTS
IN AD RISK ESTIMATION BASED ON
GENETIC ANALYSIS
In determining the risk of Alzheimer’s disease and of
other multifactorial disorders, the following two feasible
approaches have emerged:
1. The calculation of the risk based on known OR
(odds ratio) and RR (risk ratio).
2. Calculation of the risk based on the use of artificial
intelligence – neural networks.
Estimation of the risk in the context of multifactorial
diseases has been for the first time successfully
described by Yamada (2006). Yamada studied 202 polymorphisms
in 152 candidate genes related to stroke,
hypertension, metabolic syndrome, type 2 diabetes
mellitus, obesity and other diseases on 5000 mutually
unrelated Japanese individuals. Yamada described the
method of the calculation of individual risk based on
the DNA polymorphisms and life style analysis. Yamada
also showed how improvement in the patients’ life style
could reduce the risk of each studied disease.
In 2008, inspired by Yamada (2006), we introduced,
in collaboration with the Institute of Biochemistry, Faculty
of Science, Masaryk University in Brno, a method
for the prediction of Alzheimer’s disease based on the
analysis of 42 candidate gene polymorphisms in a group
of patients in Prague private clinics (Sery and Lochman,
unpublished). The results of our study have never
been explicitly published because of intervening commercial
interests and possible legal implications (the
private clinics offer DNA analyses to customers for a
fee). We analyzed ApoE, ACE, APBB2, DAPK1, KNS2,
IDE, APP, DAPK1, TOMM40 and other candidate gene
polymorphisms that were previously strongly associated
with Alzheimer’s disease. For the analysis, we used the
device BeadStation from Illumina Company, which was
then the first of its kind in the Czech Republic. In parallel
with the 42 polymorphisms in genes associated with
Alzheimer’s disease, we analyzed 384 polymorphisms in
96 persons in one run to determine risks of other multifactorial
disease such as obesity, cardiovascular diseases,
cancer etc. After analysing DNA of the first 192 people
who were interested in finding the risk of Alzheimer’s
disease, we calculated the relationship between the risks
identified by our method and the percentage and family
burden of Alzheimer’s disease estimated from the family
history of all concerned persons. We found that, using
out method, 27% of people with a family history of
Alzheimer’s disease had an increased risk of Alzheimer’s
disease and, in turn, 91% of subjects with no Alzheimer’s
disease history in the family, were identified as
having no increased risk. The relationship between the
presence of detected risk factors and family history of
Alzheimer’s disease was statistically significant at a confidence
level of p<0.02 using analysis of variance. This
study indicated that in 2008 there was enough information
available for a meaningful predictive calculation of
365Neuroendocrinology Letters Vol. 35 No. 5 2014 • Article available online: http://node.nel.edu
Alzheimer`s disease
the overall risk of Alzheimer’s disease. Since 2008, the
amount of knowledge about the relationship between
DNA polymorphisms and Alzheimer’s disease has been
exponentially increasing and from today’s perspective it
should be possible to predict the risk much more accurately
than in 2008. Our laboratory in cooperation with
Faculty of Medicine of University of Ostrava is currently
undertaking extensive research on the risk factors for
Alzheimer’s disease in a group of 1,600 persons (Povová
et al. 2013a, Povová et al. 2013b) and the evaluation of
results including a detailed analysis of the feasibility of
Alzheimer’s disease prediction can be expected in 2016.
The use of neural networks in the prediction of
disease based on the analysis of the results of genetic
analysis has not yet been exploited to an extent that it
would merit. Neural networks operate on the principle
of selecting an appropriate neural network model for
solving the problem. For example, neural networks can
extract new medical information from raw data, build
computer models that are useful for medical decisionmaking
and aid in the distribution of medical expertise
(Dayhoff & DeLeo 2001; North et al. 2003; Ohara et al.
2011). Thus a suitably selected neural network can be
fed (“taught”) a set of real data obtained from patients
and control subjects. Such selected neural networks will
then learn to differentiate subjects and distribute them
into groups of controls and patients while evaluating
and quantifying the disease risk. The advantage of this
approach compared with a calculation by OR or RR by
conventional methods is that neural networks work in
an unbiased non-prejudicial manner, independently of
researchers, analyzing the given set of data and looking
for relationships that could quite possibly escape the
attention of researchers. A diagnostic model for coronary
heart disease based on artificial neural networks
and using an array of potential genetic and non-genetic
risk factors has been described by Atkov et al. (2012).
To date, the method of neural networks has not
found much favour with molecular biologists, psychiatrists
or physicians. Possible explanation is that
the neural networks are still perceived as something
of a “black box”; using complex and seemingly nontransparent
mathematical paradigms which do not
lend themselves to easy description in methodology
sections of publications and grant applications or, for
that matter, when explaining the medical procedures
to patients and health insurance companies. Neural
networks are, however, used by bankers, finance and
market advisors as well as traders in the stock markets,
who all seem to have found common ground with science
professionals developing, exploring (and exploiting)
the artificial intelligence.
DISCUSSION
Genetic testing is currently not in routine use in psychiatry.
The chief reason is the lack of knowledge of the
pathogenetic mechanisms leading to the development
of mental disorders. Regulatory effects of epigenetic
mechanisms and the newly discovered non-coding
RNA greatly complicate the current and further research
into Alzheimer’s disease. Another problem remains the
lack of homogeneity of the groups of persons used as
human subjects in the research and the possible impact
of the environmental factors on the pathogenesis of
Alzheimer’s disease. The latest technologies make it
possible to examine associations of virtually anything
with anything else without any definite prior knowledge
of the pathogenesis of Alzheimer’s disease. On one
hand, the outcomes of such studies should be viewed
and interpreted with utmost caution when making conclusions
in relation of pathogenetic mechanisms but, on
the other hand, these new approaches may open new
possibilities in the design of predictive testing for both
Alzheimer diseases and other multifactorial disorders.
An interesting and challenging question remains –
how to use the results of predictive genetic testing for
Alzheimer’s disease at a time when we have no effective
treatment. There exists, fortunately, a degree of
knowledge of non-genetic (external, environmental)
risk factors for Alzheimer’s disease; so much so that
the persons identified at increased genetic risk of the
disease can be advised to change lifestyle or diet and
perhaps use dietary supplements. It should be noted
though that the increased risk is not a certain diagnosis
of Alzheimer’s disease. If a person has an increased
risk, changes in lifestyle could significantly delay the
disease onset by many months or years. If we look at
the matter from a population point of view, then just
one year of healthy life superimposed onto the population
of the Czech Republic – where it is estimated that
around 100,000 sufferers live with Alzheimer’s disease
– translates into 100,000 extra years of good quality life.
One cannot help but wonder what this would mean in
terms of ethical, social, and economic benefits to the
society. From this perspective we could expect that the
introduction of new DNA technologies in preventive
genetic testing for complex multifactorial diseases will
have a great positive impact.
ACKNOWLEDGEMENTS
This work has been supported by Internal grant agency
of the Ministry of Health of the Czech Republic (IGA
MZCR) No. NT/11152–6. Thanks are expressed to Dr.
Tomáš Fűrst (Olomouc) for helpful comments.
REFERENCES
1 Altar CA, Hornberger J, Shewade A, Cruz V, Garrison J, Mrazek
D (2013). Clinical validity of cytochrome P450 metabolism and
serotonin gene variants in psychiatric pharmacotherapy. Int Rev
Psychiatry. 25: 509–533.
2 Atkov OY, Gorokhova SG, Sboev AG, Generozov EV, Muraseyeva
EV, Moroshkina SY, et al (2012). Coronary heart disease diagnosis
by artificial neural networks including genetic polymorphisms
and clinical parameters. J Cardiol. 59: 190–194.
366 Copyright © 2014 Neuroendocrinology Letters ISSN 0172–780X • www.nel.edu
Omar Šerý, Jana Povová, Vladimir J. Balcar
3 Beitz JM (2014). Parkinson’s disease: a review. Front Biosci. 6:
65–74.
4 Bell JT, Tsai PC, Yang TP, Pidsley R, Nisbet J, Glass D, et al (2012).
Epigenome-wide scans identify differentially methylated regions
for age and age-related phenotypes in a healthy ageing population.
PLoS Genet. 8: e1002629.
5 Bohacek J, Gapp K, Saab BJ, Mansuy IM. Transgenerational epigenetic
effects on brain functions (2013). Biol Psychiatry. 73:
313–320.
6 Boks MP, Derks EM, Weisenberger DJ, Strengman E, Janson E,
Sommer IE, et al (2009). The relationship of DNA methylation
with age, gender and genotype in twins and healthy controls.
PLoS One. 4: e6767.
7 Chouliaras L, Ruttne BPF, Kenis G, Peerbooms O, Visser PJ, Verhey
F, et al (2010). Epigenetic regulation in the pathophysiology of
Alzheimer disease. Progress Neurobiol. 90: 498–510.
8 Collinge J (2001). Prion diseases of humans and animals: their
causes and molecular basis.  Ann Rev Neurosci. 24: 519–550.
9 Dayhoff JE1, DeLeo JM (2001). Artificial neural networks: opening
the black box. Cancer. 91(8 Suppl): 1615–1635.
10 Di Lorenzo F, Di Lorenzo B (2013). Iron and aluminum in Alzheimer‘s
disease. Neuro Endocrinol Lett. 34: 504–507.
11 Earls LR, Westmoreland JJ, Zakharenko SS (2014). Non-coding
RNA regulation of synaptic plasticity and memory: Implications
for aging. Ageing Res Rev. S1568–1637: 41–45.
12 Ferri, C.P., Prince, M., Brayne C, Brodaty H, Fratiglioni L, Ganguli M,
et al. (2005). Alzheimer’s Disease International. Global prevalence
of dementia: a Delphi consensus study. Lancet. 366: 2112–2117.
13 Hosák L (2013). New findings in the genetics of schizophrenia.
World J Psychiatry. 3: 57–61.
14 Issa JP (2014). Aging and epigenetic drift: a vicious cycle. J Clin
Invest. 124: 24–29.
15 Langley-Evans SC (2014). Nutrition in early life and the programming
of adult disease: a review. J Hum Nutr Diet. (in press)
16 Lehrner A, Bierer LM, Passarelli V, Pratchett LC, Flory JD, Bader
HN, et al (2014). Maternal PTSD associates with greater glucocorticoid
sensitivity in offspring of Holocaust survivors. Psychoneuroendocrinology.
40: 213–220.
17 Lochman J, Balcar VJ, Šťastný F, Šerý O (2013b). Preliminary
evidence for association between schizophrenia and polymorphisms
in the regulatory Regions of the ADRA2A, DRD3 and
SNAP-25 Genes. Psychiatry Res. 205: 7–12.
18 Lochman J, Plesník J, Janout V, Povová J, Míšek I, Dvořáková D, et
al (2013a). Interactive effect of MTHFR and ADRA2A gene polymorphisms
on pathogenesis of schizophrenia. Neuro Endocrinol
Lett. 34: 792–797.
19 Maloney B1, Sambamurti K, Zawia N, Lahiri DK (2012). Applying
epigenetics to Alzheimer’s disease via the latent early-life associated
regulation (LEARn) model. Curr Alzheimer Res. 9: 589–599.
20 Martín-Subero JI (2011). How epigenomics brings phenotype
into being. Pediatr Endocrinol Rev. 9 Suppl 1: 506–510.
21 Medway C, Morgan K (2014). Review: The genetics of Alzheimer’s
disease; putting flesh on the bones. Neuropathol Appl Neurobiol.
40: 97–105.
22 Miao CG, Yang YY, He X, Xu T, Huang C, Huang Y, et al (2013).
New advances of microRNAs in the pathogenesis of rheumatoid
arthritis, with a focus on the crosstalk between DNA methylation
and the microRNA machinery. Cell Signal. 25: 1118–1125.
23 Moss EG (2002). MicroRNAs: hidden in the genome. Curr Biol. 12:
R138–140.
24 Müller DJ, Chowdhury NI, Zai CC (2013). The pharmacogenetics
of antipsychotic-induced adverse events. Curr Opin Psychiatry.
26: 144–150.
25 Mychasiuk R, Harker A, Ilnytskyy S, Gibb R (2013). Paternal stress
prior to conception alters DNA methylation and behaviour of
developing rat offspring. Neuroscience. 241: 100–105.
26 North BV, Curtis D, Cassell PG, Hitman GA, Sham PC (2003).
Assessing optimal neural network architecture for identifying
disease-associated multi-marker genotypes using a permutation
test, and application to calpain 10 polymorphisms associated
with diabetes. Ann Hum Genet. 67: 348–356.
27 Ohara T, Ninomiya T, Kubo M, Hirakawa Y, Doi Y, Hata J, et al
(2011). Apolipoprotein genotype for prediction of Alzheimer‘s
disease in older Japanese: the Hisayama Study. J Am Geriatr Soc.
59: 1074–1079.
28 Oliphant A, Barker DL, Stuelpnagel JR, Chee MS (2002). BeadArray
technology: enabling an accurate, cost-effective approach to
high-throughput genotyping. Biotechniques. 2002, Suppl: 56–8:
60–61.
29 Povová J, Ambroz P, Bar M, Pavuková V, Šerý O, Tomášková H,
et al (2012). Epidemiological of and risk factors for Alzheimer’s
disease: a review. Biomed Pap Med Fac Univ Palacky Olomouc
Czech Repub. 156: 108–114.
30 Povová J, Ambroz P, Tomášková H, Bar M, Šerý O, Štěpánová M,
et al (2013a). Vascular and psychosocial risk factor of Alzheimer’s
disease. Prakticky Lekar. 93: 59–62.
31 Povová J, Tomášková H, Šerý O, Ambroz P, Vařechová K, Janout V
(2013b). Alzheimer‘s Disease and it‘s Risk for Next Generations.
Hygiena. 58: 117–120.
32 Prince M, Bryce R, Albanese E, Wimo, A, Ribeiro W, Ferri CP (2013).
The global prevalence of dementia: a systematic review and
metaanalysis. Alzheimers Dement. 9: 63–75.
33 Ronaghi M, Uhlén M, Nyrén P (1998). A sequencing method
based on real-time pyrophosphate. Science. 281: 363–365.
34 Sarantzis A, Bown MJ (2014). The genetic basis for aortic aneurysmal
disease. Heart. 100: 916–922.
35 Sarda S, Hannenhalli S. Next-Generation Sequencing and Epigenomics
Research: A Hammer in Search of Nails (2014). Genomics
Inform. 12: 2–11.
36 Šerý O, Povová J, Míšek I, Pešák L, Janout V (2013). Molecular
mechanisms of neuropathological changes in Alzheimer’s disease:
a review. Folia Neuropathol. 51: 1–9.
37 Šerý O, Přikryl R, Častulík L, Št’astný F (2010). A118G polymorphism
of OPRM1 gene is associated with schizophrenia. J Mol
Neurosci. 41: 219–222.
38 Simundic AM (2010). Methodological issues of genetic association
studies. Clin Chem Lab Med., 48 Suppl 1: S115–118.
39 Shi X, Sun M, Liu H, Yao Y, Song Y (2013). Long non-coding RNAs:
a new frontier in the study of human diseases. Cancer Lett. 339:
159–166.
40 Steemers FJ1, Gunderson KL (2007). Whole genome genotyping
technologies on the BeadArray platform. Biotechnol J. 2: 41–49.
41 Viana RJ, Nunes AF, Rodrigues CM (2012). Endoplasmic reticulum
enrollment in Alzheimer’s disease. Mol Neurobiol. 46: 522–534.
42 Wu H, Zhang Y (2014). Reversing DNA methylation: mechanisms,
genomics, and biological functions. Cell. 156: 45–68.
43 Yamada Y (2006). Identification of genetic factors and development
of genetic risk diagnosis systems for cardiovascular diseases
and stroke. Circ J. 70: 1240–1248.
44 Zhang JP, Malhotra AK (2013). Pharmacogenetics of
antipsychotics: recent progress and methodological issues.
Expert Opin Drug Metab Toxicol. 9: 183–191.
45 Zhou FC (2012). DNA methylation program during development.
Front Biol (Beijing). 7: 485–494.