A rare glimpse of schizophrenia’s genetic roots

The
delusions and hallucinations of schizophrenia can be devastating for
the 1% of the population struck by the disease. The condition clearly
has a genetic component, evidenced by its tendency to run in families.
However, the search for specific genes or chromosomal regions involved
has led to few reproducible findings.

Results published in Nature
by a multinational consortium include some of the first definitive
genetic links to schizophrenia and help to define the diversity of DNA
changes that contribute to the disease. Led by scientists from the
Stanley Center for Psychiatric Research at the Broad Institute of
Harvard and MIT, in collaboration with 10 other institutes in Europe
and the USA, the research team discovered that schizophrenic patients
were more likely to possess rare, structural changes in their genomes
than people without schizophrenia. In addition, changes at two new
sites in the genome were identified as risk factors, bringing the total
number of solid genetic links to the disorder to three and confirming
that this type of variation may help unravel schizophrenia. The study
is the largest and most complete genome-wide investigation of the
disorder to date.

Before this new work, previous genetic
association studies of schizophrenia had not produced strong,
consistent, and unambiguous findings. In 2006, a group of investigators
got together and realized that, collectively, their thousands of
samples held the statistical power to reveal genetic links that the
studies could not reveal alone. The scientists formed a new research
team known as the International Schizophrenia Consortium (ISC), which
is led by Pamela Sklar, senior associate member of the Broad Institute
and associate director of the Psychiatric and Neurodevelopmental
Genetics Unit in the Center for Human Genetics Research at
Massachusetts General Hospital (MGH). Sklar is also director of
genetics at the Stanley Center, which provided the major funding and
research resources for the current work.

The ISC is a
partnership of investigators from the University of Aberdeen, Cardiff
University, the University of Edinburgh, Karolinska Institutet, MGH,
the University of North Carolina-Chapel Hill, Queensland Institute of
Medical Research, the University of Southern California, the Stanley
Center at the Broad Institute, Trinity College Dublin, and University
College London. “With this consortium, the field of psychiatry has come
together in a way that is unprecedented,” said Sklar, “with the
understanding that no individual study has enough power to ask these
incredibly important questions.”

“The consortium should be
recognized for taking the important first step towards unearthing the
full underlying genomic architecture of schizophrenia and other
psychotic disorders,” says Edward Scolnick, director of the Stanley
Center. “Only by doing such a large study could we have uncovered these
stunning findings to such a high degree of confidence, thus setting the
stage for an even more complete understanding of the full genomic
contributions to disease.”

Cutting-edge genotyping and analysis
tools have given researchers the ability to study disease genetics on a
large scale. “In this second generation of medical genetics, we can use
new genomic technologies to look at thousands of patients at a much
finer resolution than ever before,” said Jennifer Stone, the study’s
lead postdoctoral fellow who worked closely with Sklar and Shaun
Purcell, the study’s senior analyst, on all aspects of the new work.

To
conduct this large study, the consortium researchers assayed DNA
samples totaling 3,391 individuals with schizophrenia and 3,181
unrelated individuals without the disorder. DNA samples were sent to
the Broad for genotyping and analysis, using genomic technologies and
novel analytical techniques developed at the institute and at MGH.
Taking an approach that used high-density DNA microarrays and
sophisticated analysis methods, the team looked for rare instances in
which sections of DNA are deleted or duplicated, so-called copy number
variants or CNVs, and tested whether those instances were likely to
play a role in the disease.

The team identified three deleted
regions in the genome linked to schizophrenia, including an area on
chromosome 22 observed in earlier studies. The other two deletions,
though, had not been previously implicated in the disease. Large
sections of DNA on chromosomes 1 and 15 were missing more often in
patients with schizophrenia than in controls. Moreover, in a study
appearing in the same Nature
issue, a team of researchers led by scientists at deCODE Genetics found
that the same three deletions conferred risk for the disorder,
signifying a major step forward in schizophrenia genetics. “It is
encouraging that both research teams identified these same genomic
sites,” said Broad associate member Shaun Purcell, who is an assistant
professor of psychiatry at Harvard Medical School and senior analyst on
the new study.

In addition to identifying distinct genetic
risk factors, the team suspected that the overall level of rare copy
number changes — both extra and missing DNA — might differ between
cases as a group and controls as a group. This “CNV burden” can be
measured either by counting the total number of missing or extra DNA
segments in the genome or by counting the number of genes those
segments intersected.

Both measures of burden were found to
differ — cases showed a 1.15-fold higher rate of CNVs than controls,
and the “gene count” burden was even higher at 1.41-fold. As genes are
the major functional units of DNA, it is perhaps not surprising that
CNVs that alter genes have a more pronounced effect versus those that
do not.

“In light of these new findings, there no longer needs
to be an argument that this type of variation plays a role in some of
the risk,” said Sklar. “The task now is to understand how rare
structural variation, together with other kinds of genetic variation
and environmental effects, work together to produce schizophrenia.”

Sklar
adds that the work was made possible by the consortium’s collaborative
efforts and by the resources and funding provided by the Stanley Center
and the Broad Institute. The new findings are some of the first
successes to come out of the Stanley Center, established in early 2007
through a $100 million grant from the Stanley Medical Research
Institute. The center was created to combine the power of genomics and
chemical biology in an effort to improve the understanding, diagnosis,
and treatment of bipolar disorder, schizophrenia, and major depression.
“This wonderful framework has enabled the development of new tools and
made them rapidly available,” Sklar said.

Also critical to the
success of this unprecedented study are the analytical tools developed
by Shaun Purcell and other researchers at the Broad and MGH, such as
PLINK, an analytical tool for associating genomic data with disease
that is now used worldwide by genomic researchers. “PLINK is not only
crucial to our studies of psychiatric genetics, but because it and
other analysis tools are made publicly available, they are becoming the
international standard for whole-genome studies,” said Sklar.

Other
scientists in the Broad community, especially lead research technician
Kimberly Chambert and members of the Broad Institute’s research
platforms, also contributed greatly to the success of this
high-quality, complicated, and large-scale genetics project, said
Sklar. In particular, members of the Biological Samples Platform, led
by Kristin Ardlie, and the Genetic Analysis Platform, led by Stacey
Gabriel, helped to rapidly produce the massive amount of data required
for this study.

In total, more than 6,000 large, rare copy
number variants were identified. Because rare deletions and
duplications were also found in controls, albeit at a lower frequency
than in cases, the genetic changes that are truly risk factors have yet
to be clearly identified. “Just observing one of these events in a
patient does not in itself tell you anything about the role of that
particular variant in causing that person’s disease,” said Purcell.
Further studies will be necessary to understand which variants are
disease-causing and, in larger variants that contain many genes, which
specific genes are disease-related.

The team will continue to
analyze the consortium’s data. The microarrays used to analyze the
sample also allow for measurement of common variation in DNA sequence
and common copy number variation. Along with the data on rare
structural variation, the forthcoming analysis of common variation may
help complete the picture of genetic variation in schizophrenia. Sklar
calls this the $64,000 question: What is the mix of all the different
kinds of variation in the genome of patients and how do they contribute
to the risk of disease?

“With these tools in hand,” said
postdoctoral fellow Stone, “we now have a strong foothold into the
genetic architecture of this psychiatric disease.” Skolnick adds that
while these new genetic links are indeed remarkable, they are truly the
first steps towards revealing how genetic variation contributes to the
risk of developing schizophrenia. Additional work over the next several
years will be needed to fully expose and understand the genetic
underpinnings of this complex disease.

The International
Schizophrenia Consortium is part of the larger Psychiatric Genome-wide
Association Study Consortium, which seeks to rapidly analyze the tens
of billions of genotypes generated for five central psychiatric
disorders — autism, attention-deficit hyperactivity disorder, bipolar
disorder, major depressive disorder, and schizophrenia. In addition to
enabling large-scale studies of each disorder individually, consortium
members will be searching for risk factors that span multiple diseases,
potentially revealing unique perspectives into the genetics of
psychiatric diseases as a whole.

Other Broad researchers
participating in the work include Douglas Ruderfer, Joshua Korn, Mark
Daly, Steve McCarroll, Casey Gates, and Scott Mahon.