Why do you study genome variation?
Neurodegenerative diseases are very
common diseases of the elderly popu-
lation and there is currently no effec-
tive therapy or cure available. Genetic
mutations play an important role in
the risk for developing neurodegener-
ative diseases such as Alzheimer’s and
Parkinson’s disease. Mutations that
are biologically damaging can have
such a strong effect that they cause
a disease without other influencers.
Although they are usually rare in the
general population you find them in
a few families where the disease fol-
lows a Mendelian inheritance pattern.
Other mutations have a much small-
er effect and by themselves are not
strong enough to cause a disease and
they can be quite frequent in the pop-
ulation because there is no evolution-
ary selection against them. However,
if an individual carries several of these
risk factors and is exposed to environ-
mental risk factors as well, the combi-
nation could also lead to disease.
We try to identify genetic risk factors
because they give us a unique handle
to study the mechanisms that result
in a disease. The mutations are the di-
rect cause of the disease and we carry
genetic mutations with us our whole
life from the moment of conception,
therefore they function at the very
beginning of the disease process. The
mutations are therefore an ideal start-
ing point for research that aims to un-
derstand the mechanisms that lead to
a disease and to start to work towards
developing a therapy to fix or prevent
the problem.
How exactly do you study genetic
risk factors?
In genetics we look for genome vari-
ants that are shared between pa-
tients, but not shared with healthy
individuals. There are many ways to
look at DNA and to study variations in
our genome but nowadays the most
widely used techniques are DNA se-
quencing and genotyping.
The field of genetics has benefitted
greatly from the technological devel-
opments that are the direct result of
the Human Genome project. The goal
of the Human Genome project was to
sequence a complete human genome
but the technology to do this efficiently
20 | NEUROMAG | July 2018
did not exist at the start of the project
in 1990. It took 10 years to complete
the first draft of the human genome.
Now, we can sequence a full genome
in a matter of days. This allows us to
compare the complete genomes of
patients and healthy individuals and
identify the differences. The costs for
whole genome sequencing are still
very high but there are good alterna-
tives that work well for a fraction of
the costs. Based on the variation that
has been found in all the sequencing
efforts on the human population we
have collected 100 thousands vari-
ants that have been spotted on glass
arrays that we can use to find regions
of the genome shared much more fre-
quently by patients than by healthy
individuals. With these arrays we can
now test thousands of individuals to
search for genetic risk variants using
what we call Genome Wide Associa-
tion Studies (GWAS).
How is it possible to link genome
mutations with a disease?
For Mendelian diseases, where a sin-
gle mutation is enough to cause dis-
ease, we search for variants that are
present in patients but not in healthy
individuals. Often we have families
available with multiple affected indi-
viduals and the variants should co-
segregate with the disease. Once we
have a list of these variants we use
bioinformatics approaches to deter-
mine which of these variants are likely
to have a damaging effect. Finally, we
will test the identified variants in a cel-
lular model system to confirm their
damaging effect and study the mech-
anism by which they could cause the
disease.
For weaker risk factors the situation
is more complicated because a single
risk factor is not sufficient to cause
the disease and therefore also many
healthy people in the population can
carry some of the risk variants. We
therefore have to use statistical meth-
ods and large numbers of patients and
healthy individuals to find variants
that are significantly more present
in disease cases versus healthy con-
trols. A second complication is that
since whole genome sequencing on
thousands of samples is still very, very
expensive and we have to use geno-
typing arrays. With these arrays we
can find a region where the risk vari-
ant is located using GWAS but we can’t
detect the variant itself. In fact these
regions often contain multiple genes
and it remains difficult to prove which
gene and risk variant is responsible for
the increased risk.
Our lab uses a series of approaches to
resolve this problem. We can sequence
the region identified by GWAS in our
study samples and then use bioinfor-
matics to identify those variants that
are most likely to have a functional
effect. Unfortunately, the knowledge
of all the functions encoded in the hu-
man genome is still limited, especially
for the noncoding part of the genome
which is more the 98% of our genome.
We work tirelessly to improve this by
participating in international consortia
such as the FANTOM consortium that
has helped to find many new noncod-
ing RNA genes, microRNAs and regu-
latory elements, such as enhancers,
and we use this information to help
find the most likely gene or variant
that increases the risk for a given dis-
ease.