Kein Folientitel

1
Identification and functional characterisation
of molecular risk factors in acute
leukemias Renate Kirschner1, Michaela Heide1,
Peter Rhein1, Leonid Karawajew1, Matthias
Nees2, Stephen Breit2, Andreas Kulozik2,
Christian Hagemeier1, Wolf-Dieter Ludwig1,
Rainer Spang3, Karl Seeger1 1 Charité,
Humboldt-Universität Berlin, 2 Universitäts-Kinder
klinik Heidelberg, 3 Max-Planck-Institut für
molekulare Genetik Berlin,
Childhood acute lymphoblastic leukemia (ALL)
occurs with an incidence of approximately 600
patients per year in Germany. In general, up to
75 of children can be cured permanently by
chemotherapy. ALL relapses (approximately 100
cases per year) are more resistant to treatment
with a cure rate of less than 50. Therefore,
novel approaches in terms of diagnosis and
therapy are particularly needful for this group.
In order to identify novel prognostic factors and
to unravel molecular mechanisms underlying
clinical outcome, we aim to generate gene
expression profiles of initial and relapsed ALL
of the Berlin-Frankfurt Münster (ALL-BFM and
ALL-REZ BFM) study group by Affymetrix? DNA
microarray technology.
Bone marrow samples for gene expression profiling
are selected from patients who have been treated
according to the protocols of the ALL-BFM and
ALL-REZ BFM study groups for intial and relapsed
acute lymphoblastic leukemia, respectively. For
a retrospective study patients are classified by
clinical outcome, whereas for a prognostic study
proven risk factors as for example minimal
residual disease - MRD will be used to divide
patients into subgroups. MRD sensitively measures
the amount of leukemic cells that are still
present at certain time points during therapy.
Optimising Gene Expression Profiling

• RNA Preparation

Isolation of Minor Subpopulations from
Heterogeneous Leukemic Samples
In the second part of the project, we focus on
distinct, clinically relevant subpopulations from
initially heterogeneous leukemic cell samples. We
are especially interested on minor subpopulations
of immature, progenitor-like leukemic cells as
well as on residual leukemic cell populations
which have escaped initial treatment and become
more resistant to therapy. In order to approach
this issue experimentally, procedures for
identification and purification of rare leukemic
blast cells based on flow cytometric analysis and
flow sorting are under development.
In order to perform retrospective studies, we
isolated RNA from cryopreserved mononuclear cells
(magenta). In a significant number of samples
loss of sufficient RNA quality and quantity was
observed. RNA quality was also evaluated on a
Bioanalyzer and via hybridisation of Affymetrix
Test3Arrays showing a high degree of
consistency. For prospective studies we therefore
routinely prepare RNA directly from incoming bone
marrow biopsies with an optimized yet straight
forward protocol (light blue). This should also
minimize changes of expression profiles due to
cryopreservation. Retrospective studies can only
be performed on a limited number of samples
(lt25).

• Evaluation and cross-validation of generated
  data sets with
• published ALL expression profiles

Databanks were created containing published genes
found to be deregulated in large scale ALL
profiling studies based on Affymetrix U95Av2
GeneChips (eg Yeoh et al. 2002. Cancer Cell 1
133-143). Ongoing co-hybridisation of 5 to 10
Probes to U95Av2 and U133A GeneChips will allow
normalization of expression profiles for
comparing data sets created with old and new
Affymetrix GeneChips. Databanks and bioinformatic
filtering tools can then be used to identify and
select against unwanted signatures in smaller
test populations that would otherwise escape
recognition. This part therefore aims at both,
utilizing and cross-validating existing data from
different laboratories.