European Human Genetics Conference 2007 June 16 – 19, 2007 ...
European Human Genetics Conference 2007 June 16 – 19, 2007 ...
European Human Genetics Conference 2007 June 16 – 19, 2007 ...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Concurrent Sessions<br />
been replicated in numerous studies in different populations. However,<br />
evidence for the underlying mechanism how USF1 variants relate to<br />
dyslipidemia remains limited.<br />
We here show for the first time, allele-specific differences in USF1<br />
transcript levels in two relevant tissues. In muscle biopsies from 142<br />
twins, following insulin challenge only subjects homozygous for the<br />
non-risk allele responded with significantly increased USF1 transcript<br />
levels, as measured by RT-PCR. Essentially the same phenomenon<br />
was observed in fat-biopsies of FCHL family members as insulin levels<br />
correlated with allelic imbalance. Quantitative sequencing of genomic-<br />
and transcribed RNA from subjects heterozygous for the best associating<br />
SNP, located in the 3’-UTR of USF1 revealed an average ~20%<br />
lower expression of the risk-allele in fat.<br />
In a set of 47 expression arrays of fat-biopsies, differential expression<br />
of numerous USF1 target genes was evident -among them many<br />
genes of lipid metabolism and inflammatory-response. Interestingly,<br />
the neighboring F11R gene recently implicated in development of atherosclerotic<br />
plaques also showed allele-dependent expression, suggesting<br />
the presence of multiple, potentially co-regulated CVD associated<br />
genes in this chromosomal region.<br />
In summary, common allelic variants, defined by non-coding SNPs of<br />
the USF1 transcription factor gene and associated with dyslipidemia<br />
risk, seemingly eradicate the insulin response of transcript levels. Subsequently,<br />
they result in differential expression of numerous relevant<br />
target genes predisposing carriers of risk alleles to dyslipidemia and<br />
CVD.<br />
C66. The cohesion protein NIPBL recruits histone deacetylases<br />
to mediate chromatin remodeling<br />
W. Xu1 , P. Jahnke1 , M. Wülling2 , M. Albrecht1 , G. Gillessen-Kaesbach1 , F. J.<br />
Kaiser1 ;<br />
1 2 Universitätklinikum S-H,, Lübeck, Germany, Universität Duisburg-Essen, Essen,<br />
Germany.<br />
Cornelia de Lange Syndrome (CdLS) is a rare malformation disorder<br />
with multiple congenital anomalies, a characteristic face, growth and<br />
mental retardation as well as gastrointestinal and<br />
limb abnormalities. About 50 % of the patients with CdLS carry mutations<br />
in the NIPBL gene.<br />
NIPBL encodes a homologue of the fungal Scc2-type and Drosophila<br />
Nipped-B protein and is part of the chromatid cohesion complex.<br />
Recent studies show an association of chromatid cohesion with chromatin-remodeling<br />
complexes, either in the recruitment of cohesion to<br />
particular sequences along chromosome arms or in the establishment<br />
of cohesion.<br />
In yeast-two hybrid assays we could identify the chromatin remodeling<br />
factors histone deacetylases 1 and 3 (HDAC1 and 3) as potential NIP-<br />
BL-interacting proteins. Using different fragments of NIPBL in liquid βgalactosidase<br />
assays, we could narrow down the interacting region for<br />
HDAC1 and 3 to a stretch of <strong>16</strong>2 amino acids (aa) within a highly conserved<br />
region of NIPBL which was predicted to be a HDAC-interacting<br />
domain by in silico studies. In luciferase reporter gene assays we could<br />
show that this HDAC-interacting domain of NIPBL fused to the GAL4-<br />
DNA-binding domain (GAL4-DBD) is able to repress reporter gene<br />
transcription. Moreover, cotransfections of HDAC1 and 3 enhance this<br />
effect significantly. To confirm whether this effect is based on histone<br />
deacetylation we used an antibody recognizing the acetylated histone<br />
3 in chromatin immunoprecipitation assays and could identify specific<br />
deacetylation of histone 3 adjacent to the NIPBL-GAL4-DBD binding<br />
sites. Our data show that NIPBL is able to recruit chromatin remodeling<br />
enzymes mediating histone deacetylation.<br />
C67. The <strong>Human</strong> Epigenome Project (HEP)<br />
S. Beck;<br />
Wellcome Trust Sanger Institute, Cambridge, United Kingdom.<br />
Epigenetic processes play an essential role in biology with wide-ranging<br />
implications for human health and disease. To understand and harness<br />
these processes we need to read and interpret the epigenetic<br />
code with the same rigour and vigour that made reading the genetic<br />
code one of the greatest scientific achievements. To this end, a number<br />
of efforts have already been initiated of which the EU-funded HEP<br />
was among the first to be set up in 2000 with the aim to map one of the<br />
epigenetic marks, DNA methylation. On behalf of the HEP Consortium,<br />
I will present our findings to date, discuss some of the lessons learnt<br />
and give an outlook on how the data and technology may be used in<br />
an integrated (epi)genetic approach to common disease.<br />
www.epigenome.org<br />
C68. Mechanism of Alu integration into the human genome<br />
J. Chen 1,2 , C. Férec 2,3 , D. N. Cooper 4 ;<br />
1 Etablissement Français du Sang<strong>–</strong>Bretagne, Brest, France, 2 INSERM, U613,<br />
Brest, France, 3 Université de Bretagne Occidentale, Faculté de Médecine de<br />
Brest et des Sciences de la Santé, Brest, France, 4 Institute of Medical <strong>Genetics</strong>,<br />
Cardiff University, Cardiff, United Kingdom.<br />
LINE-1 or L1 has driven the generation of at least 10% of the human<br />
genome by mobilising Alu sequences. Although there is no doubt that<br />
Alu insertion is initiated by L1-dependent target site-primed reverse<br />
transcription, the mechanism by which the newly synthesised 3’ end<br />
of a given Alu cDNA attaches to the target genomic DNA is less well<br />
understood. Intrigued by observations made on 28 pathological simple<br />
Alu insertions, we have sought to ascertain whether microhomologies<br />
could have played a role in the integration of shorter Alu sequences<br />
into the human genome. A meta-analysis of the <strong>16</strong>24 Alu insertion<br />
polymorphisms deposited in the Database of Retrotransposon Insertion<br />
Polymorphisms in <strong>Human</strong>s (dbRIP), when considered together<br />
with a re-evaluation of the mechanism underlying how the three previously<br />
annotated large deletion-associated short pathological Alu inserts<br />
were generated, enabled us to present a unifying model for Alu<br />
insertion into the human genome. Since Alu elements are comparatively<br />
short, L1 RT is usually able to complete nascent Alu cDNA strand<br />
synthesis leading to the generation of full-length Alu inserts. However,<br />
the synthesis of the nascent Alu cDNA strand may be terminated prematurely<br />
if its 3’ end anneals to the 3’ terminal of the top strand’s 5’<br />
overhang by means of microhomology-mediated mispairing, an event<br />
which would often lead to the formation of significantly truncated Alu<br />
inserts. Furthermore, the nascent Alu cDNA strand may be ‘hijacked’<br />
to patch existing double strand breaks located in the top-strand’s upstream<br />
regions, leading to the generation of large genomic deletions.<br />
C69. Evolution of mammalian gene expression promoted by<br />
retroposons<br />
N. V. Tomilin;<br />
Russian Academy of Sciences, St.Petersburg, Russian Federation.<br />
Expression of eukaryotic protein-coding genes is strongly affected by<br />
specific sequence motifs which concentrate near transcription start<br />
sites (TSS) and bind transcription factors. These transcription factor<br />
binding motifs (TFBM) may arise in promoters by slow accumulation<br />
of base changes in a random sequence but recent data indicate that<br />
promoter TFBMs can also arise rapidly through insertion of mobile elements<br />
already having functional motifs or having sequences which<br />
may be converted to TFBMs by a small number of mutations. For example,<br />
single C to T transition at specific position can create functional<br />
estrogene-response element in human Alu retroposons. We studied<br />
abundance of retroposon-derived sequences in human and mouse<br />
promoters (-1000 to +200 bp segments relative to TSS) and found that<br />
>15000 of human promoters and >11000 of mouse promoters have<br />
Alu-derived or B1/B2-derived elements, resp. Global distribution of<br />
these retroposons in human and mouse chromosomes strongly correlates<br />
with clusters of CpG islands present in promoters of ~75% genes.<br />
In active genes CpG islands are not methylated but major fraction of<br />
retroposons, especially LINEs and LTRs, are heavily methylated.<br />
It is unknown how promoter CpG islands are protected from spread<br />
of methylation initiated at adjacent LINEs and LTRs. <strong>Human</strong> Alu and<br />
mouse B1 elements can be transcribed by RNA polymerase III because<br />
their internal promoters bind transcription complex TFIIIC. This<br />
complex can limit spread of repressive histone methylation in S.pombe<br />
and we suggest that Alu- and B1-associated TFIIIC sites have similar<br />
function in mammals.<br />
C70. Prominent use of distal 5’ transcription start sites and<br />
discovery of a large number of additional exons in ENCODE<br />
regions<br />
A. Reymond 1 , F. Denoeud 2 , J. Harrow 3 , C. Ucla 4 , A. Frankish 3 , C. Henrichsen 1 ,<br />
C. Wyss 4 , T. Alioto 2 , J. Drenkow 5 , J. Lagarde 2 , T. Hubbard 3 , S. E. Antonarakis 4 ,<br />
R. Guigo 2 , P. Kapranov 5 , T. R. Gingeras 5 ;<br />
1 Center for Integrative Genomics, University of Lausanne, Lausanne, Switzerland,<br />
2 IMIM, Barcelona, Spain, 3 Wellcome Trust Sanger Institute, Hinxton,<br />
1