Functional Kinase Activity Biomarkers in clinical translational research
A biomarker is a substance and a measurable indicator of normal biological processes, pathogenic processes, or response to therapy. A good biomarker can contribute to a reliable diagnosis, tell us what therapy is the best for a particular patient, or even better: tell us whether a person is at risk for getting a certain disease before getting to the stage where it has already occurred and need treatment. In breast cancer the ER, PR and HER2 expression have implications for prognosis and therapy selection that are independent of TNM staging. For melanoma, there are few predictive or prognostic markers available. Only recently have mutations within BRAF at codon 600 (BRAF(V600E)) been validated as a marker with implications in prognosis and treatment. However, although better markers are identified and used for diagnostic and prognostic purposes with implications in therapy treatment, histological examination is still required for diagnosis, whereas immunohistochemical and genetic tests are utilized for treatment decisions and prognosis determinations. The value of current diagnostic and prognostic markers has been fairly weak, and the mechanisms leading to cancer progression are still poorly understood. In this project we identified possible regulatory and functional genomic biomarkers in breast cancer and melanoma through high throughput screening methods. We used both tumor tissue and in vitro cell line models to perform molecular analysis that could provide us with good biomarkers for personalized medicine. The focus has been on markers that regulate gene expression (i.e. regulatory biomarkers) such as microRNAs (miRNAs) and also functional biomarkers such as protein kinases that are often deregulated in cancer, including breast cancer and melanoma. For breast cancer, we have characterised the miRNA expression and kinase activity in relation to known molecular, clinical and histopathological parameters. The studies performed in breast cancer demonstrated that miRNAs can be used as biological markers in all stages of the disease, discriminating between normal, benign and malignant tissue. Some miRNAs that were previously identified as cancer-related miRNAs were identified to be deregulated in benign tumors aswell, suggesting that miRNA deregulation might be seen already at early stages of tumor formation, making them good candidates for early tumor detection. Furthermore, analysis of miRNA expression and tyrosine kinase activity in a subset of breast cancer exhibited differential miRNA expression and tyrosine kinase activity in breast tumors that were ER+ and HER2-, with and without PR expression, making them good markers for identifying tumor subtypes. Lacking PR expression exhibited a high activity of tyrosine kinases, especially kinases involved in immunity pathways. Next, we analyzed gene expression profiles in matched tumor and normal breast tissue compared to normal, disease free breasts. Age and BMI were identified as important contributors to differences obtained in gene expression in adjacent normal tissue of women with breast cancer. Normal tissue adjacent to tumors was genetically different from tissue not affected by tumors, especially for macrophage- associated markers. In melanoma, we analyzed the kinase activity in tumor tissue and normal tissue, revealing that nearly all kinases were highly active in tumors. Ex-vivo treatment with vemurafenib exhibited differential inhibition of tyrosine kinases within MAPK and PI3K pathway in BRAF(V600E) and BRAF-wild type malignant melanoma patients, unravelling possible novel targets for therapy within these pathways. All in all, the use of microarray technology has provided valuable information regarding the molecular composition of different types of tumors, identifying possible biomarkers that can be useful in the diagnosis, prognosis and treatment of breast cancer and melanoma in the future.
I dette prosjektet er det blitt oppdaget nye markører for brystkreft og melanom som kan eventuelt hjelpe til med diagnose, prognose eller behandling av disse sykdommene. Dette er blitt gjort ved å kartlegge molekylære funksjoner i tumorceller og normale celler. Fokuset har vært på regulatoriske biomarkører i form av små molekyler som mikroRNA (miRNA) som regulerer genuttrykk i cellene. Videre har det også blitt forsket på funksjonelle biomarkører som proteinkinaser, kjent for å være deregulert i kreft og er ypperlige kandidater til målrettet behandling. Forskning på brystkreft viste at flere miRNA kunne separere mellom normalt, godartet og ondartet vev. Sammen med data fra genekspresjon og proteinkinaseaktivitet, kan disse markørene brukes til å identifisere forskjellig tumortyper i brystkreft. Proteinkinaser viste seg å være av stor betydning også i melanom, spesielt for behandlingen av den typen som omfatter den spesifikke mutasjon BRAF(V600E). Alt i alt, ved å bruke molekylære teknikker er det blitt bidratt til å avdekke nye potensielle biomarkører og frembrakt nyttig informasjon relatert til den molekylære komposisjonen til forskjellige grupper av brystkreft og melanom. Selv om resultatene er lovende, må det mer forskning til før disse markørene kan brukes i klinikken.
Tyrosine Kinase Activity Profiling and gene expression profiling in subsets of breast cancer patients
The study has focused on gene and miRNA expression profiles in breast cancer and normal tissue, and also protein tyrosine kinase activity in breast cancer subtypes. We have studied the molecular composition of different types of breast cancer samples in order to identify biomarkers that can be related to diagnosis, prognosis or treatment of breast cancer patients.
Kinases are considered as promising source of biomarkers for prognostic and therapeutic purposes in cancer patients. They are usually found to be dergulated in cancer and their use in targeted therapy including kinase inhbitiors have increased during the past decades. In this project we assessed tyrosine kinase activity in 32 primary breast cancer samples that were all hormone receptor positive (ER+ and/or PR+) with differential HER-2 status, using microarray technology. We aimed to get a better understanding of the complexity of the molecular biology in subsets of breast cancer patients. We used Pamchip® peptide microarrays to measure the activity of protein tyrosine kinases in 32 breast cancer samples, and also breast cancer cell lines MCF-7, BT474 and ZR75-1. The cell lines were studied for kinase activity both untreated and treated with estradiol. Results showed differences in phosphorylation amongst breast cancer samples. A total of 37 peptide kinases were highly phosphorylated in a group of breast cancer samples representing 33 protein tyrosine kinases involved in cancer pathways and immunological responses. In vitro studies with breast cancer cell lines exhibited the same phosphorylation profiles, but increased phosphorylation was only observed in one cell line, ZR75-1. By eliminating HER-2 positive samples, we obtained differences in phosphorylation profiles based on PR-status only. Samples lacking PR-expression exhibited higher kinase activity of downstream kinases compared to PR-positive samples. Similar results were obtained with miRNA expression profiles of 31 of the same breast cancer samples. Five miRNAs were identified to be significantly differentially expressed (p < 0.05) between PR-negative and PR-positive samples. This effect was even stronger when eliminating HER-2 positive samples from the analysis, with 13 miRNAs exhibiting significant differential expression based on PR-status. Although our data are based on a small dataset, the lack of PR expression seems to have a profound effect on tyrosine kinase activity and miRNA expression in HER-2 negative breast cancers without any effect on gene expression. This indicates that regulatory and functional molecules might exhibit phenotypical features of cancer that cannot be explained by gene expression alone. In the second study we compared gene expression in breast cancer tissue and their corresponding adjacent normal tissue in relation to covariates such as tumor subtype, patient age and body mass index (BMI). Gene expression profiling was performed in 195 breast tumors and their corresponding matched adjacent normal tissue, as well as 43 normal tissue samples from women undergone mammoplastic reduction without cancer. Age and BMI were independently associated with inflammation in normal tissue but not tumors, with increased expression of macrophages and other inflammatory cells seen in patients with high BMI (>30) and older age (>50). Additionally, macrophage expression was lower in ER-positive tumors compared to adjacent normal tissue in direct proportion to increased expression of ESR1 and genes with ESR1-binding activity, suggesting elevated estrogen levels may affect tumor-resident macrophages and immune cells in these patients. We calculated a novel statistic to quantify the rewiring of gene co-expression networks and demonstrated that basal gene networks are rewired in ER-positive tumors, even though their gene expression levels are not significantly different from the adjacent normal tissue. This study illustrates how patient age and tumor subtype interact to affect inflammation and proliferation in breast tumors and the tumor microenvironment.
Functional Kinase Activity Biomarkers in Translational Research
Within this project we have performed tyrosine kinase activity arrays on melanoma samples with and without the V600E mutation, and studied the effect of the drug directed against this mutation. Profound differences in kinase activity were observed that could predict drug response dependent on the presense/absence of the V600E mutation.
In the present project period the studies re extended to perform tyrosine kinase arrays with breast cancer samples. Firstly, we are comparing breast cancer samples that are estrogen receptor positive, and either HER2-negative or HER2 –positive. The tyrosine kinase arrays contain 144 peptides representative of the downstream targets of tyrosine kinases present in the sample. The ER-positive and HER2-positive tumors are usually resistant towards endocrine therapy, and it is important to elucidate differences between these two tumor types, and what pathways ae up- or downregulated. The study is validated with treated cell lines representing the patient samples. The PamGene’s technology platform developed by the Dutch biotechnology company, PamGene International, is established at EpiGen and is further utilised to to examine the activity of the interaction of nuclear hormone receptors with coregulators in tissue obtained from strictly defined therapy settings at Ahus and OUS. The performance and interpretation of the experiments demand for experts from different fields: surgeons (to collect the samples), clinical oncologists (for the clinical and treatments response follow-up) molecular biologists (for identifying the relevant biological pathways), biostatisticians for data analysis and technical assistants to ensure the continuous use of the instrument that is described. The PamChip technology for NHR arrays is relatively new, and allows us to investigate the broad spectrum of interactions of nuclear receptors, in our case ER-alpha, with 154 coregulator peptides present on the PamChip® through PamGene's Microarray Assay for Real-time Coregulator-Nuclear Receptor Interaction (MARCoNI). The PamChip® disposable consists of 12 identical arrays with each array containing up to 154 immobilized nuclear receptor coregulator peptides harboring either LXXLL (coactivator) or LXXXIXXXL (corepressor) motifs. These peptides are immobilized in a porous microarray membrane, and the peptide-binding of the nuclear receptor in the assay mix is visualized through fluorescence antibodies captured by a camera, and detected in a PamStation®12 processing platform (https://www.pamgene.com/upload/image/brochures/NHR-Pamgene2014-PRINT_4.pdf). This platform will provide an opportunity for the institution to utilize large biobanks and associated clinical databases compiled over several years. Additionally, the platform will be able to provide a unique, centralized, cost-effective laboratory facility to serve both important clinical environments at OUS, Ahus and other hospitals of the South East region in agreement with the strategic plan of Helse SØ. This will represent an important counterbalance towards the well established core facilities in the Oslo University Hospital and will, in collaboration with the other core facilities, provide the necessary free capacity to accommodate projects from the South-Eastern Norway Health Region. This technology has so far been available only on a service basis from the PamGene International B.V. Transferring the technology to Norway will contribute to generate knowledge and competence in the field of targeted therapies. This is final work in the drgrad, which is planned to be delivered dupring spring.
Functional Kinase Activity Biomarkers in clinical translational research
A platform for the studies of kinase activity at EpiGen is used to study the differential inhibition of ex vivo tumor kinase activity by vemurafenib in BRAF(V600E) and BRAF wild-type metastatic malignant melanoma. Ex-vivo exposure of tumor samples to kinase inhibiting drugs may reveal mechanisms of resistance and identifying predictive markers.
Recently we established a platform for the studies of kinase activity at EpiGen. Several related clinically relevant projects wëre initiated with our clinical collaborators (Jurgen Geisler, Ida Bukholm, Anne Hansen Ree). The PhD student Adlena Tahiri published 2 papers in the first 2 years of the project. On maternity leave, will continue from June, plans to deliver end of 2014.
In this project we aimed at a broader characterization at the kinome in samples obtained from patients suffering from metastatic melanomas using a multiplex peptide array technology. Kinase activity profiling was performed using the Tyrosine Kinase PamChip® Array for Pamstation®12 (PamGene International B.V., ‘s-Hertogenbosch, The Netherlands) at Akershus University Hospital. Each array consists of 144 peptide substrates, primarily with known tyrosine residues, representing ~100 different proteins. Three chips can be run simultaneously, and each chip consists of four arrays. The lysates are repeatedly pumped up and down through the porous array, allowing repeat substrate phosphorylation. The samples were run in two experimental series. In the first experiment, the samples were run in three technical replicates in the presence and absence of vemurafenib, as paired measurements with and without inhibitor on the same chip. The measurements were repeated with sunitinib. Based on pilot experiments of increasing concentrations of the individual kinase inhibitors added to melanoma tissue lysates, concentrations that resulted in ~50% inhibition of most array substrates were chosen for the patient sample experiments. Concentrations of 40 µM vemurafenib and 7.5 µM sunitinib were spiked into the assay mixtures prior to incubation, whereas 1.5% dimethyl sulfoxide to mixtures not containing the inhibitor. Incubations were commenced for 60 cycles, followed by washing and fluorescence measurement of all peptide spots every fifth cycle. The experiments were run blinded, and the tumor and normal skin tissue lysates were run separately. The microarray data are submitted to ArrayExpress (http://www.ebi.ac.uk/arrayexpress/); accession number E-MTAB-1245. Our findings suggest that this technique may contribute significantly to the better understanding of the network of kinases involved in metastatic melanomas in general. and may identify interacting tumor signaling pathways that can result into the development of alternative actionable therapy targets for metastatic disease control in malignant melanoma. In addition, ex-vivo screening of drugs may help to identify subsets of kinases as predictive markers and subsets of kinases to improve patient selection at bed-side for targeting treatments like BRAF-inhibitors (Tahiri et al, PLos1, 2013).
Our next goal will be to characterise miRNAs that may be associated to the observed tirosine kinase profiles. We have recently observed a strong association between the expression of a number of miRNAs and the levels of a number of protein, many of which kinases (Aure et al, unpublished). We wouls like to charachterise the miRNA profiles in the same melanoma samples from paper 1. The candidate is experiences with miRNA analysis. She perfromed recently an extensive miRNA study "Deregulation of cancer-related miRNAs is a common event in both benign and malignant human breast tumors" (Tahiri et al, Carcinogenesis, 2014)
Kinase Activity Biomarkers in translational research
The knowledge that the oncogene BRAF is frequently mutated in human melanoma (40-50% of cases) has caused the development of BRAF-targeting kinase inhibitors, like vemurafenib (PLX4032) and dabrafenib. Better understanding of the molecular mechanisms involved in response to vemurafenib therapy is urgently needed.
Purpose: In this study, we assessed the individual kinase activity profiles in tumor samples obtained from patients with metastatic malignant melanoma. In addition, we studied the overall ex-vivo inhibitory effects of vemurafenib and sunitinib on kinase activity status.
Experimental design: Metastatic tumor samples obtained from twenty-six melanoma patients were analyzed using peptide microarrays with kinase substrates. Individual kinase activity profiles were generated from tissue lysates before and during ex-vivo incubation with the kinase inhibitors vemurafenib and sunitinib.
Results: Overall kinase activity was significantly higher in lysates from metastatic melanoma tumors compared to normal skin tissue. Ex-vivo incubation with vemurafenib or sunitinib caused significant decrease in phosphorylation of kinase substrates. While basal phosphorylation profiles were similar in BRAF wild-type and BRAF(V600E) tumors, we identified a subset of 40 kinase substrates that were more strongly inhibited by BRAF(V600E) tumor lysates incubated with vemurafenib, and could distinguish between BRAF wild-type and BRAF(V600E) tumors. Interestingly, a few BRAF wild-type tumors showed inhibition profiles similar to BRAF(V600E) tumors.
Conclusion: Multiplex kinase substrate arrays may give valuable information about overall tumor kinase activity. Ex-vivo exposition of tumor samples to kinase inhibitors may be useful to study mechanisms of therapy resistance and to identify predictive markers.
Tyrosine kinase profiling in metastatic malignant melanoma
Identification of possible therapeutic targets and prediction markers for response to therapy
First manuscript under preparation
Metastatic melanoma is an aggressive form of skin cancer that is highly resistant to conventional chemotherapy. Thus, novel approaches to improve the current treatment of advanced melanoma are urgently needed. Kinases are key regulators in the cell, and since deregulation is implied in cancer, the information of kinase activity in tumors give important tumor-specific information on affected pathways and potential targets for treatment. With this in mind, we performed tyrosine kinase profiling of 26 tissue samples obtained from patients suffering from stage IV melanoma, prior to dacarbazine treatment. The samples were equally divided between responders and non-responders. The technology used was based on a unique 3D flow-through microarray format, ideal for working out kinase pathways. Each Tyrosine Kinase PamChip® Array contained 144 peptide substrates with sites for phosphorylation, representing 100 different proteins. During incubation, images were taken during each cycle, allowing the monitoring of the reaction in real time. Using the array, each sample lysate generated an individual phospho-substrate signature, representing the state of information flow through kinase signaling cascades. The technology is useful for many purposes as it allows for functional comparison of biological samples without any prior knowledge to which signalling pathways are activated. Previous research on kinases has led to the discovery of multiple receptor kinase inhibitors used as therapeutics for patients with advanced cancers. Sunitinib acts on RTKs such as VEGFRs, PDGFRs and KIT, known to be involved in tumor development and progression. In addition, B-RAF inhibitor, like PLX4032 (Vemurafenib), has shown dramatic effects in advanced melanoma tumors, and has recently been approved for treatment against late stage melanoma. However, the drug is only beneficial for melanoma patients with B-RAF V600E mutations, and resistance to the drug through activation of alternative survival pathways is a still a problem. Consequently, incubation was carried out on all samples with and without kinase inhibitors (Sunitinib and PLX4032). The use of kinase inhibitors will cause silencing of corresponding subsets of activated signaling pathways; hence, providing us leads to therapeutic targeting of the regulatory mechanisms underlying cellular processes. Our preliminary results showed that PLX4032 acted on multiple tyrosine kinases, even though the compound is mainly directed against a serine/threonine kinase, B-RAF. The detailed effects of both PLX4032 and sunitinib in the chosen test system are currently prepared for publication.
Tracing the origin of disseminated tumor cells in breast cancer using single-cell sequencing.
Genome Biol 2016 Dec 09;17(1):250. Epub 2016 des 9
Determinants of acquired activated protein C resistance and D-dimer in breast cancer.
Thromb Res 2016 Sep;145():78-83. Epub 2016 aug 3
LIMT is a novel metastasis inhibiting lncRNA suppressed by EGF and downregulated in aggressive breast cancer.
EMBO Mol Med 2016 Sep 01;8(9):1052-64. Epub 2016 sep 1
Identification of a Natural Killer Cell Receptor Allele That Prolongs Survival of Cytomegalovirus-Positive Glioblastoma Patients.
Cancer Res 2016 Sep 15;76(18):5326-36. Epub 2016 jul 12
Subtype-specific micro-RNA expression signatures in breast cancer progression.
Int J Cancer 2016 Sep 01;139(5):1117-28. Epub 2016 mai 9
Integrated analysis reveals microRNA networks coordinately expressed with key proteins in breast cancer.
Genome Med 2015;7(1):21. Epub 2015 feb 2
Glycan-related gene expression signatures in breast cancer subtypes; relation to survival.
Mol Oncol 2015 Apr;9(4):861-76. Epub 2015 jan 14
Deregulation of cancer-related miRNAs is a common event in both benign and malignant human breast tumors.
Carcinogenesis 2014 Jan;35(1):76-85. Epub 2013 okt 8
PMID: 24104550 - Inngår i doktorgradsavhandlingen
Systematic assessment of prognostic gene signatures for breast cancer shows distinct influence of time and ER status.
BMC Cancer 2014;14():211. Epub 2014 mar 19
Differential inhibition of ex-vivo tumor kinase activity by vemurafenib in BRAF(V600E) and BRAF wild-type metastatic malignant melanoma.
PLoS One 2013;8(8):e72692. Epub 2013 aug 30
PMID: 24023633 - Inngår i doktorgradsavhandlingen
Gefitinib in Combination with Weekly Docetaxel in Patients with Metastatic Breast Cancer Caused Unexpected Toxicity: Results from a Randomized Phase II Clinical Trial.
ISRN Oncol 2012;2012():176789. Epub 2012 mai 14
Tyrosine Kinase Activity Profiling Reveals Differences According to Progesterone-Receptor-Status in HER2 Negative Breast Cancer
San Antonio Breast Cancer Symposium 2015
Age and estrogen-dependent inflammation in breast adenocarcinoma and normal breast tissue
submitted to Molecular Oncology
Differential kinase inhibition in BRAF(V600E) positive metastatic malignant melanoma by vemurafenib
under 2 revision, inngår i avhandlingen
Tyrosine kinase profiling in metastatic malignant melanoma: Identification of possible therapeutic targets and prediction marker
Annual meeting American Association for Cancer Research, Chicago 2012