A kutatási téma leírása:
I. BACKGROUND, PREVIOUS WORKS AND OBJECTIVES
Systemic chemotherapies are widely used for the treatment of breast cancer; they can reduce the risk of distant metastases by approximately one-third. Multiple combinations of cytotoxic drugs are applied empirically despite the observation that the available regimens are not equally effective across the population of patients with a particular type of cancer. Predictive markers are factors that are associated with response or resistance to a particular therapy. We have investigated several markers and tested their predictive value in breast cancer (Surowiak, Gyorffy et al, Brit J Cancer, 2006). We have also investigated the molecular mechanisms responsible for drug resistance (Abdul-Ghani, Gyorffy et al, Oncogene, 2006). However, to date, – besides estrogen receptor – no single tumor marker has been validated to possess a sufficient predictive value to render it clinically useful (Gyorffy et al, Cancer Genomics and Proteomics, 2005) (Gyorffy, Recent Results in Cancer Research, 2007). As drug resistance or response almost certainly depends on the interplay of multiple genes, the investigation of large gene-sets – using high-density gene expression microarrays – should produce reliable predictive tests.
In recent years a large number of microarray experiments were performed, often yielding contradictory results. van’t Veer et al. identified gene expression signatures predictive for short interval to distant metastases. Interestingly, none of the individual genes previously described in the literature were present in their predictive gene set of 70 genes. Several papers investigated the response to doxorubicin, an anthracycline drug applied for breast cancer, GI cancer, lung cancer, cervix carcinoma and more. Kang et al. (2004) identified gene expression patterns related to resistance against 5-fluorouracil, cisplatin and doxorubicin resistance in fourteen human gastric cancer cell lines. Hess et al. (2006) published a 30-probe set pharmacogenomic predictor predicting pathologic complete response (pCR) to preoperative weekly paclitaxel and fluorouracil-doxorubicin-cyclophosphamide chemotherapy with high sensitivity. Gianni et al. (2005) found 24 genes correlating with pCR in patients with locally advanced breast cancer receiving neoadjuvant paclitaxel and doxorubicin. Cleator et al. (2005) published a response prediction gene list containing 30 genes predicting the response of breast tumors to neoadjuvant adriamycin and cyclophosphamide chemotherapy. Folgueira et al. (2005) found a trio of genes capable to potentially distinguish doxorubicin-responsive from nonresponsive tumors. Ayers et al. (2004) identified a predictive gene list for the sequential paclitaxel and fluorouracil +doxorubicin +cyclophosphamide protocol in breast cancer.
We used gene expression patterns of well-defined chemotherapy-resistant and sensitive cancer cell lines to identify discriminatory genes associated with drug resistance and applied our model to predict sensitivity in a set of 44 previously characterized breast cancer samples. The patient group characterized by the gene expression profile similar to that of sensitive cell lines exhibited longer survival (49.77±26.1 months, n=21, p=0.034) than the resistant group (32.97±18.7 months, n=23). Thus, the application of gene expression signatures derived from doxorubicin-resistant and -sensitive cell lines allowed to predict effectively clinical survival after doxorubicin monotherapy. Our approach demonstrated the significance of in vitro experiments in the development of new strategies for cancer response prediction (Gyorffy et al, Oncogene, 2005). In another study we tested 30 cancer cell lines for sensitivity to 5-fluorouracil, cisplatin, cyclophosphamide, doxorubicin, etoposide, methotrexate, mitomycin C, mitoxantron, paclitaxel, topotecan and vinblastine at drug concentrations that can be systemically achieved in patients. The resistance index was determined to designate the cell lines as sensitive or resistant, and then, the subset of resistant vs. sensitive cell lines for each drug was compared. An individual prediction profile for the resistance against each chemotherapy agent was constructed, containing 42–297 genes. The overall accuracy of the predictions in a leave-one-out cross validation was 86%. A list of the top 67 multidrug resistance candidate genes that were associated with the resistance against at least 4 anticancer agents was identified. Moreover, the differential expressions of 46 selected genes were also measured by quantitative RT-PCR using a TaqMan micro fluidic card system (Gyorffy et al, Int J Cancer, 2006).
In the above gene expression signatures all together five hundred genes were associated with doxorubicin resistance. We have compared all these publications and the published sets of differentially regulated genes with each other, and we have found only 5% overlap in the gene lists (Gyorffy et al, Anticancer Res, 2006). Similar results were obtained in another paper describing gene lists associated with melanoma tumorigenesis (Gyorffy et al, J Invest Dermatol, 2007), 5-fluorouracil resistance (Szoke, Gyorffy et al, Onkologie, 2007) and ovarian tumorigenesis (Gyorffy et al, Int J Gynecol Cancer, 2008). Thus, the repeated measurements generally do not confirm the results of previous publications. Why do we have different gene lists when comparing different microarray experiments for the same clinical question? The most likely reason behind this phenomenon is the fact, that most genes with altered expression represent downstream targets of upstream regulatory elements. As some of the tests are already in diagnostic use (like the van’t Veer gene set) it is imperative to spot these real causative genes behind the altered expression signatures.
However, functional validation is usually accomplished by a gene-by-gene basis, creating a bottle-neck effect for the characterization for the huge numbers of targets arising from genomic surveys. To solve these issues I suggest the development of transfected cell arrays (TCA, the method will be discussed in the next chapter).
The project will enable to find genes whose altered expression correlates with chemotherapy response. While the identified genes will provide a new generation of targets for drug development, the major aim and the most significant impact of our study will be the establishment of a new technology workflow to measure the role of genes uncovered in different microarray experiments. While today we are lacking any functional hypothesis which might explain the reason why a given set of genes is co-regulated, the project will enable to discriminate regulatory elements and downstream targets.
II. METHODS AND PRELIMINARY RESULTS
Over the last several years RNA interference (RNAi) has emerged as a powerful approach for manipulating gene expression in mammalian cells. During RNAi, the introduction of double-stranded RNA into a cell decreases the level of the complementary mRNAs producing a knockdown of the corresponding protein. The current model of the RNAi mechanism proposes that the silencing trigger is processed by Dicer into small RNAs of 21–22 nucleotides in length (Bernstein et al, 2001). These become incorporated into an RNA-induced silencing complex with endonuclease activity, which, in turn, identifies and cleaves homologous mRNAs (ZamorE et al, Nat Struct Biol, 2001, Hannon, Nature, 2002). The most promising development in this field is the RNAi cell microarray that makes possible discrete, in-parallel transfection with thousands of RNAi reagents on a single microarray slide. In this method, nucleic acids are printed on a standard glass microarray slide and used to transfect cells with the help of a transfection reagent. The addition of cultured cells in medium to the array forms a living cell microarray of locally transfected cells in a lawn of nontransfected cells. The nucleic acid species (plasmids, linear PCR products or oligonucleotides) detach or desorbs from the solid phase and are taken up by the cells, the uptaken DNA is overexpressed within the cells or, when using RNAi, the target gene is silenced. The technology is also called “reverse transfection” because cells are added on top of the nucleic acid instead of vice versa (Ziauddin and Sabatini, Nature, 2001, Mishina et al. J Biomol Screen, 2004). Depending on the transfection competency and the size of the cells being used, each feature of the array can consist of 30-500 transfected cells. The largest published RNAi cell microarray to date were performed in cultured D. melanogaster cells covering 384 genes and using high-resolution automated image acquisition and analysis (Wheeler et al., Nat Methods, 2004).
In present study the aim is to focus on breast cancer. Concerning the application of microarray technology, breast cancer is one of the most promising fields where already a large number of studies were published. In particular, the focus will be on doxorubicin resistance, a drug included in 3 out of 6 protocols for breast cancer chemotherapy. One might consider focusing one of the newer targeted therapy agents, like trastuzumab, bevacizumab, cetuximab, as resistance against these agents is also very diverse (for trastuzumab see Diermeier et al, Exp Cell Res, 2005, Nahta et al, Cancer Res, 2004, Lu et al, J Natl Cancer Inst, 2001, Clark et al, Mol Cancer Ther, 2002), and individual response to these drugs is difficult to predict. However, we have selected doxorubicin because 1, we have our own resistant cell lines (as published previously – see literature) and 2, there are several published genome-wide gene expression patterns which make current experiment feasible. The established workflow can then be easily applied for other therapy agents.
While establishing the methods (cell culture, transfection and positive controls), the first step of the study will be the identification of genes to be included in the study and the construction of the glass arrays containing these genes. For the selection of genes we will use our own data and publicly available datasets. The oligos for the selected genes will be picked from commercially available RNAi libraries to achieve fast and reproducible results. Despite rules that have been established to choose effective siRNAs, unknown secondary mRNA structure or other factors render many RNAi constructs unable to initiate RNAi. Therefore, we will use five different oligos to silence a single gene. The DNA chip spotter we used for the construction of spotted DNA microarrays will be used to construct the cell arrays. The application of TCA has many advantages compared to conventional RNAi: it is estimated that between 100 and 500 reverse transfections can be done with the materials required for a single transfection in a well of a 96-well plate (Silva et al, PNAS, 2004). TCA also requires fewer cells per number of genes tested (Vanhecke and Janitz, Oncogene, 2004). Finally, using TCAs the screening of very large sets of genes will not require significant automation, and for small arrays (on the order of 100 features), additional multiplexing can be achieved by simply printing duplicate arrays on the bottoms of the slides to give ‘arrays of arrays’.
The nucleic acid oligos will be resuspended in DNA-condensation buffer and a transfection reagent will be added. Then, the mix will be added to 0.2% gelatin to complete a transfection master mix. After the slides are briefly exposed to a lipid transfection reagent, human breast cancer cell lines MCF7 and MDA231 and their doxorubicin-resistant derivatives will be added on top of the slides and cells landing on the printed areas become transfected with the printed oligo. After 48-72 hours of culture without media change the cells will be treated with doxorubicin. While genes up-regulated in resistance will be investigated in the resistant cell lines, down-regulated genes will be applied to the parental cell line to achieve resistance. To identify RNAs whose expression affects chemoresistance, the live cell clusters will be visualized 24h after drug treatment and cell protein content will be measured after fixation using the sulphorhodamine-B assay.
We must consider an important limitation of the TCA technology, which has to be addressed during our analyses. Low (<10%) transfection efficiencies have been reported in the solid-phase as opposed to in-solution transfection (Delehanty et al. Anal Chem 2004, Baghdoyan et al. Nucleic Acids Res 2004). However, a minimal number of transfected cells per cluster will be required to attain a statistically accurate sampling. Therefore, different transfection reagents will be used to assist transfection in TCAs. High solid-phase transfection efficiencies might also be achieved upon addition of fibronectin (Yoshikawa et al. J Control Release, 2004). Moreover, the use of the previously designated cell lines might not be feasible. In this case we will switch to easily transfectable cell lines, such as human embryonic kidney cells (HEK293T) (Silva et al, PNAS, 2004) and monkey kidney fibroblasts (COS) (Bailey et al. Drug Discov Today 2002). Should all these approaches still not solve the problem, we will consider the use of electrostatic control of DNA adsorption and desorption, which has been extensively studied and has recently been shown to improve transfection efficiency (Hook, et al Biosens Bioelectron. 2006).
Another important question will be the use of an appropriate marker gene. Green fluorescent protein is suggested as a reporter protein for plasmid uptake and intracellular expression (Webb et al. J Biomol Screen, 2003). The application of RNAi on printed slides inhibiting the expression of a co-delivered marker gene has also been demonstrated (Mousses et al. Genome Res. 2003, Kumar et al, Genome Res 2003). In present project, preference will be on fluorescence detection due the possibility to collect signal intensities from the whole slide using standard microarray scanners.
In a recent test experiment we prepared siRNA-gelatin transfection solution in 384-well plates. siRNA solution were mixed with Effectene®, and incubated at room temperature with gelatin and fibronectin. The solutions were arrayed onto poly-L-lysin covered glass slides using a solid pins resulting in a spot diameter of approximately 400 μm for all experiments. The spot volume was approximately 4 nL. After printing, 100.000 NIH3T3 cells were plated on the Petri dish containing one glass slide and incubated for 44 h. However, these results were achieved using a selected easily transfectable cell line and already validated genes. As transfection optimization is expected to be a time-intensive task, an appropriate time is allocated in the time-plan. For the evaluation of the TCA results we will apply the Cellprofiler software (http://www.cellprofiler.org; Wheeler et al, Nature Genetics, 2005).
The third step will be the validation of the significantly changed genes selected in the cell-array experiment. For this purpose, routine 6-well RNAi will be performed and the gene expression changes will be monitored using genome-wide gene expression microarrays. The gene expression changes will be validated using RT-PCR.
The final stage will include the application of the selected genes for clinical patient therapy-response prediction. As our approach will enable the selection of key genes associated with resistance we expect to have a high prediction potential using these genes as response predictors. Due to limited time and to save resources, we do not aim to build a new tumor bank for this project. Rather we will utilize publicly available datasets with known therapy protocols and response data. Thus, this will be an in silico application of results.
The statistical evaluations will be performed in the R statistical environment. The significance of individual genes will be calculated using the significance analysis of microarrays as described by Tusher et al (2003). Previously published microarray results will be annotated and matched using the Microsoft Access software and the Affymetrix Netaffx analysis centre. Patient survival and response prediction will be performed using a semi-supervised principal component method as described by Bair et al (2004). The statistical significance will be set to have a false discovery rate below 5%.
The results will be published in international scientific journals and international cancer conferences. As the results will provide validated candidate genes for drug resistance, these genes will be patented for future drug developments (special attention will be given to separate the scientific publications and patent issues).
felvehető hallgatók száma: 3
Jelentkezési határidő: 2019-01-30