Development of a procedure for genome wide expression profiling from minute Tissue samples and application in mammary carcinoma [Elektronische Ressource] : gene activity patterns unveiling molecular pathways and predicting clinical response / presented by Olaf Thürigen

Dot

Sujets

• Gesundheit

Informations

• Publié par: ruprecht-karls-universitat_heidelberg
• Publié le: 01 janvier 2008
• Nombre de lectures: 18
• Langue: English
• Poids de l'ouvrage: 3 Mo

Extrait

Inaugural Dissertation

Submitted to the
Combined Faculties for the Natural Sciences and for
Mathematics

of the
Ruperto-Carola University of Heidelberg,
Germany

for the degree of
Doctor of Natural Sciences



Presented by
Dipl.-Biol. Olaf Thürigen
Born in Ludwigshafen
Oral Examination: January 30, 2008ii

Development of a Procedure for
Genome-wide Expression Profiling from
Minute Tissue Samples and
Application in Mammary Carcinoma:
Gene Activity Patterns Unveiling Molecular
Pathways and Predicting Clinical Response













Referees:
Prof. Dr. Werner Buselmaier
Prof. Dr. Peter Lichter
iii
iv




















ʺThe harder you work, the luckier you get.ʺ 

Gary Player
vAcknowledgements (in order of appearance)
- My family, especially my parents, for building and being my solid
foundation,
- Jörg Schlingemann for introduction to Peter, our fruitful collaboration
succeeding in TAcKLE, and good friendship,
- Peter Lichter for giving me the opportunity to earn my doctorate
and his persevering attitude in the slopes of this (or any) endeavor,
- Meinhard Hahn, for schooling my eye for the detail
and using his own copiously,
- Gunnar Wrobel and Felix Kokocinski for profound introduction into "R"
and valuable help with management of large data,
- Heidi Schlingemann (then Kramer) for great help and good fun in the
laboratory,
- Grischa Tödt for his broad and sophisticated (bio)informatics deployment,
concerning both our collaborative work and home cinema systems,
- Björn Tews for his superb introduction into RQ-PCR
and an example in sportsmanship,
- Daniel Haag for supporting my work and his wonderful cocktail recipes,
- Andreas Schneeweiß and Hans-Peter Sinn for providing excellent material
and data, resulting in successful collaborations,
- Benedikt Brors and Patrick Warnat for help in identifying relevant and
predictive genes and pathways,
- Magdalena Schlotter for patiently answering my newer ending
"Bambi questions",
- Prof. Dr. Werner Buselmaier, for taking my supervision, being my first
referee and the chairperson of the examining committee,
- All my former and current colleagues in the expression lab, in the Dept. of
Molecular Genetics, and of course all my friends at the DKFZ, and
- Astrid Riehl for discussion of this thesis, and our future.

vi α. Acknowledgements vi
β. Contents vii
γ. Summary ix
δ. Zusammenfassung x
ε. Abbreviations xi

1. Introduction 1
1. Cancer 1
1. Cancer Development
2. Cancer Progression Models
2. Breast Cancer 4
1. Breast Cancer Types
2. Hereditary Mammary Carcinoma
3. Clinical Treatment of Breast Cancer
4. Diagnosis and Treatment Options
5. Molecular Profiling in Prognosis and Therapy Response Prediction
3. DNA Microarrays 20
4. Messenger RNA Amplification Methods 24

2. Aim and Procedure 27

3. Material and Methods 28
1. Microarrays 28
1. Generation of Microarrays
2. Hybridization and Post-Processing
3. Scanning and Data Pre-Proc
2. Messenger RNA Amplification and Labeling Protocol 34
1. Sample and Reference RNA
2. Comparative Amplification and Labeling of RNA
3. Analysis of Comparative Amplifications
3. Gene Expression Signature Predictive for Chemotherapy in Primary Breast
Cancer 53
1. Patients and Chemotherapy Protocol
2. Tumor Samples and Reference
3. Amplification of mRNA from Samples and Reference RNA
4. Hybridization to Microarrays and Data Pre-Processing
5. Data Analysis
6. Identification of the Gene Expression Signature
4. Real-time Quantitative PCR 60
1. Reverse Transcription
2. Primer Design
3. RQ-PCR Measurements and Analysis
5. Antibody Generation 64
1. Preparation of cDNA and Cloning into Expression Vectors
vii2. Protein Expression in the Prokaryotic System and Isolation
3. Immunization of Mice and Generation of Hybridoma Cells
4. Validation by Western Blotting
6. Pathway Analysis 70
7. Immuno-Histochemistry 71

4. Results 73
1. Messenger RNA Amplification and Labeling Protocol 73
1. Incorporation of Fluorescently Labeled Dyes
2. Performance of Amplified Messenger RNA on Oligonucleotide Microarrays
2. Gene Expression Signature Predictive for Chemotherapy in Primary Breast
Cancer 85
1. Identification of the Gene Expression Signature
2. Genes and Pathways of the Predictive Signature
3. Validation of Microarray Results by Real-time Quantitative PCR 93
4. Generation of Antibody using Mouse Hybridoma Cells 96
5. Immuno-histochemical Analysis of Patients 98

5. Discussion 103
1. Messenger RNA Amplification and Labeling Protocol 103
2. Gene Expression Signature Predictive for Chemotherapy in Primary Breast
Cancer 107
3. Genes and Pathways involved in Prediction of Chemotherapy Response 113
4. Antibody Generation using Mouse Hybridoma Cells 127
5. Immuno-histochemical Analysis of Tumors 128

6. Outlook 131

7. References xiv

8. Appendix xxiv
A. Cell Lines xxiv
B. Manufacturers of Chemicals and Laboratory Material xxv
C. Primers for RQ-PCR xxvi
D. Genes Contained in the Predictive Gene Expression Signature xxvii
E. Publications and Patent xl

viii Summary
In this thesis, a novel procedure for linear amplification of messenger RNA
(mRNA) molecules and labeling with fluorescently modified nucleotides was
developed, that can be used to perform genome-wide expression analysis from
minute tissue samples using microarrays of long gene-specific oligonucleotide
DNA probes. The procedure was then applied to analyze core needle biopsies
taken at time of diagnosis from tumors of female primary breast carcinoma
patients. Upon receiving chemotherapy consisting of gemcitabine, epirubicin
and docetaxel, the patients were classified according to their response to the
chemotherapy into responders, defined as patients with a pathological
complete remission of the tumor, and non-responders, defined as patients with
no change or pathological partial remission.
The gene expression profiles of the tumors from these patients were then
bioinformatically processed and analyzed to identify a gene expression
signature, which could be used to predict the response of the patients.
Additionally, this gene signature was inspected for the significantly enriched
pathways and biological processes, and a subset of genes was analyzed in the
patient's biopsies with respect to RNA expression as validated by real-time
quantitative polymerase chain reaction and protein expression as measured by
immuno-histochemistry.
The gene expression signature contained 512 genes, which allow a prediction of
the patient response with an overall accuracy of 88%, a sensitivity of 78% and
a specificity of 90%. Signaling pathways and biological processes identified
with significant enrichment in the gene set were the Ras pathway, TGF β
signaling, DNA damage response and apoptosis. From these pathways, the
genes DAPK2, BAMBI, LMO4 and SMAD3 could be validated by RQ-PCR, but
not SRC. In protein analysis by IHC, BAMBI was strongly associated with the
patient's outcome, while BMP4, LMO4, SMAD3 and SRC were not directly
associated. Additionally, BAMBI protein expression showed strong relationship
with BRCA1 expression in the primary female breast carcinoma.
Taken together, these results show the applicability of the novel developed
procedure for amplification and labeling of mRNA for genome-wide gene
expression analysis with the long oligonucleotide microarray technique and the
successful use in biological and clinical investigations. The analysis of gene
expression profiles of the primary breast tumors revealed an association of the
Ras pathway, TGF β signaling, DNA damage response and apoptosis with the
outcome of the patients after chemotherapy, as well as associations of several
genes within these pathways and biological processes.
ixZusammenfassung
In der vorliegenden Dissertation wurde eine neue Methode zur Amplifikation
von Boten-RNS und der Markierung mit fluoreszierenden Nukleotiden
entwickelt, die sich zur Erstellung von Genexpressions-Profilen aus sehr
kleinen Gewebeproben mit Hilfe genspezifischer Proben aus langen DNS-
Oligonukleotiden auf Microarrays eignet. Nachfolgend wurde diese Methode
angewendet, um Feinnadel-Biopsien aus Mamma-Karzinomen zu untersuchen,
die den Patientinnen bei Diagnose entnommen worden waren. Nachdem die
Patientinnen eine Kombinations-Chemotherapie aus Gemcitabin, Epirubicin
und Docetaxel erhalten hatten, wurden sie je nach Ansprechen in "Responder",
definiert als Patientinnen mit pathologisch gesicherter kompletter Remission,
oder "Non-Responder", definiert als Patientinnen ohne Veränderung oder mit
partialem Rück