2009 Vienna - European Society of Human Genetics

Concurrent Symposia
invariable in the healthy populations, but increased in these cancerprone
individuals (p=0.01). Expanding on the initial analysis and using
multiple algorithms and a replication cohort of LFS DNA (n=22) hybridized
on Affymetrix 6.0GW high resolution arrays we now demonstrate
that: 1) CNV deletions are more frequent than duplications in mutation
carriers (p=3.28 x 10 5 ); 2) the overall number of CNVs exceeds 300
in carriers (154 in controls); and 3) the difference in CNVs persists in
these individuals’ tumors. We have also recently uncovered rare CNV
deletions at 17p13.1 in LFS patients (n=4) and in patients with multiple
congenital abnormalities but no cancer phenotype (n=4). This led us
to posit that structural features of specific CNVs in the region may
play a role in distinguishing a cancer from a non-cancer phenotype.
The CNVs at 17p13.1 all include TP53 but range in size from 4.236
to 2Mbp. The large deletions, causing happloinsufficiency of nearly 40
genes, are associated with a broad phenotype that includes multiple
congenital abnormalities, while smaller focal TP53 deletions (from exons
2 to 11) are associated with cancer susceptibility. Using qPCR and
long-range PCR, we continue to fine map these deletions with the goal
of finding the sequence features and mechanisms which cause these
large genomic events (e.g. LCR-mediated NAHR). Our results suggest
an important role of DNA structural variation in cancer susceptibility,
and these models will be discussed during this presentation.
s08.3
Familial gastric cancer
C. Caldas;
Cancer Research UK, Cambridge Research Institute, Dep.of Oncology, Functional
Breast Cancer Genomics Lab, Cambridge, United Kingdom.
No abstract received as per date of printing. Please see www.eshg.
org/eshg2009 for updates.
s09.1
molecular karyotyping: From postnatal to preimplantation
genetic diagnosis?
J. Vermeesch;
Department of Human Genetics, Afdeling CME-UZ, Leuven, Belgium.
Molecular karyotyping or genome wide array CGH has been implemented
in postnatal diagnosis of patients with idiopathic mental retardation
and congenital anomalies and is challenging conventional
karyotyping as the prime diagnostic tool. Despite its successes, interpretation
of the results coming from arrays with ever increasing resolution
is becoming the main challenge. I will demonstrate how “Mendelian
copy number variants” - apparently benign CNVs that can cause a
disease phenotype dependent on copy number state, sex and genetic
or environmental background - require large scale collaborative efforts
to collect sufficient data and the development of expert systems to
provide accurate diagnosis. The technology has, more recently, been
applied in a prenatal diagnostic setting. I will illustrate how the technology
helps prenatal diagnosis, but also demonstrate the potential risks
of using this technology. Finally, we developed a novel tool to genome
wide screen CNV and SNP-genotype single cells. When applied to
cleavage stage embryos from young fertile couples we discovered,
unexpectedly, an extremely high incidence of chromosomal instability,
a hallmark of tumorigenesis. Not only mosaicisms for whole chromosome
aneuploidies and uniparental disomies but also frequent segmental
deletions, duplications and amplifications that were reciprocal
in sister blastomeres were detected in most cleavage stage embryos
implying the occurrence of breakage-fusion-bridge cycles. As a consequence,
PGD-AS will not improve the selection of genetically normal
embryos. This not only explains the low human fecundity but also
identifies postzygotic chromosomal instability as a leading cause of
constitutional chromosomal disorders.
s09.2
Prenatal diagnosis and fetal treatment using fetal RNA in
maternal body fluids
D. Bianchi;
Tufts Medical Center, Genetics, Boston, MA, United States.
Cell-free fetal (cff) DNA in maternal plasma provides a noninvasive
source of fetal genetic material. Cff DNA is elevated in preeclampsia,
placental abnormalities, and fetal aneuploidy. Qualitative analysis of
cff DNA in maternal plasma is already in clinical use worldwide for
noninvasive prenatal diagnosis of Rhesus D and fetal gender. Excit-
ing work has recently been published that suggests that cff DNA and
RNA in maternal plasma can facilitate noninvasive prenatal diagnosis
of trisomies 18 and 21, using ratios of single nucleotide polymorphisms
(SNPs) or using a shotgun sequencing approach. In our laboratory
we are performing comparison gene expression microarray analyses
between the pregnant woman and her newborn to detect fetal gene
sequences that are indicative of normal and abnormal fetal development
in the second and third trimesters (Maron et al. J Clin Invest
2007; 117:3007-3019). Amniotic fluid supernatant is also a rich source
of cell-free fetal DNA and RNA; it can provide novel information about
gene expression and functional development in the living human fetus.
We have generated preliminary data on fetal gene expression from
cell free mRNA in amniotic fluid. This has led to the identification of
novel differentially-regulated genes, and disrupted biologic pathways
in various fetal pathologies such as twin to twin transfusion syndrome,
fetal hydrops, and trisomy 21. Functional genomic analysis of second
trimester fetuses with trisomy 21 suggests that oxidative stress, ion
transport, and G-protein signaling are important. Most recently, we
have used the Connectivity Map to generate testable hypotheses regarding
fetal treatment with small molecule drugs. The Connectivity
Map “connects” disease states, biological systems disruption (as measured
by pathway analysis), and pharmaceutical compounds to treat
the disease. This allows a true translation from bench to bedside, and
suggests a possible continuum between prenatal diagnosis and fetal
therapy. Supported by National Institutes of Health R01 HD42053-06.
s09.3
the challenge of prenatal and preimplantation genetic diagnosis
of mitochondrial disorders
C. de Die-Smulders, H. Smeets;
University Hospital Maastricht, Department of Clinical Genetics, Maastricht, The
Netherlands.
Mitochondrial diseases are caused by defects in the oxidative phosphorylation.
They can be caused by mutations in nuclear or mitochondrial
DNA (mtDNA) encoded genes. Mutation analysis of nuclear
genes in prenatal (PND) or preimplantation genetic diagnosis (PGD) is
routine. Pitfalls do occur in biochemical analysis in PND. Mutations in
the mtDNA lead to a wide spectrum of diseases with very variable clinical
expression. They are transmitted exclusively maternally and are
usually heteroplasmic. Severity of symptoms is partially determined by
mutation load. PND or PGD for mtDNA mutations is complex and experience
is limited. Prerequisites for reliable PND of mtDNA mutations
have been formulated by Poulton and Turnbull (2000) and include: 1.
a close correlation between mutation load and disease manifestation
2. no significant time-dependent changes in mutation load 3. a uniform
distribution of mutation load in different tissues. These criteria also apply
for PGD.
For genetic counseling one may subdivide the mitochondrial mutations
in 5 categories: 1 de novo mutations. Recurrence risk is low and
PND or PGD can be offered for reassurance. 2. Inherited stable mutations,
such as the m.8993T>G/C mutations (leading to NARP/Leigh
syndrome). Outcome is favourable for this mutation when the mutation
load is < 60%, while mutation load >90% is associated with a bad
prognosis. Prediction of severity in the grey zone (60-90%) is difficult,
but a tendency for percentages at the extremes has been observed
in oocytes, which would favour conclusive results. PND was offered
more than 10 times. PGD for the m.8993T>G mutation has been reported
once. 3. Inherited unstable mutations. The classical example
is the m.3243A>G (MELAS) mutation. There is a certain correlation
between mutation load and clinical severity but individual exceptions
exist. It is impossible to define a completely safe lower threshold. A
limited number of PNDs have been carried out. Mutant load was found
to be fairly stable in CVS and amniotic cells. Oocytes and foetuses of
carrier women can be without mutant load, the number dependent on
the mutation load of the carrier. PGD for the 3243A>G mutation was
carried out by our group (unpublished results). Mutation load showed
a broad range between embryos, but was equal in the blastocysts of
one embryo. Some embryos had a fairly low mutation load. 4. Rare
mutations with unknown outcome. Insufficient information is available
for reliable predictions. 5. Homoplastic mutations. PND or PGD is useless
as 100% of the mtDNA is mutated.
In conclusion, assessment of mtDNA mutation load in chorionic villi,
amniotic cells or blastocysts is quite accurate nowadays and prelimi-