Florian T. Unger

Surgical procedures and postsurgical tissue processing significantly affect expression of genes and EGFR-pathway proteins in colorectal cancer tissue.

An understanding of tissue data variability in relation to processing techniques during and postsurgery would be desirable when testing surgical specimens for clinical diagnostics, drug development, or identification of predictive biomarkers. Specimens of normal and colorectal cancer (CRC) tissues removed during colon and liver resection surgery were obtained at the beginning of surgery and postsurgically, tissue was fixed at 10, 20, and 45 minutes. Specimens were analyzed from 50 patients with primary CRC and 43 with intrahepatic metastasis of CRC using a whole genome gene expression array. Additionally, we focused on the epidermal growth factor receptor pathway and quantified proteins and their phosphorylation status in relation to tissue processing timepoints. Gene and protein expression data obtained from colorectal and liver specimens were influenced by tissue handling during surgery and by postsurgical processing time. To obtain reliable expression data, tissue processing for research and diagnostic purposes needs to be highly standardized.

Metabolomics profiling of pre-and post-anesthesia plasma samples of colorectal patients obtained via Ficoll separation

Abstract1H NMR spectroscopy was used to investigate the metabolic consequences of general anesthesia in the plasma of two groups of patients with diagnosis for non-metastatic colorectal cancer and metastatic colorectal cancer with liver-metastasis, respectively. Patients were treated with etomidate or propofol, two frequently used sedation agents. Plasma samples were obtained via Ficoll separation. Here, we demonstrated that this procedure introduces a number of limitations for NMR-based metabolomics studies, due to the appearance of spurious signals. Nevertheless, the comparison of the 1H NMR metabolomic profiles of patients treated with etomidate or propofol at equipotent dose ranges was still feasible and proved that both agents significantly decrease the plasma levels of several NMR-detectable metabolites. Consequently, samples collected during anesthesia are not suitable for metabolic profiling studies aimed at patient stratification, because interpersonal variability are reduced by the overall depression of metabolites levels. On the other hand, this study showed that plasma metabolomics could represent a valuable tool to monitor the effect of different sedation agents and/or the individual metabolic response to anesthesia, providing hints for an appropriate tuning of personalized sedation procedures. In our reference groups, the metabolomic signatures were slightly different in patients anesthetized with etomidate versus propofol. The importance of standardized collection procedures and availability of exhaustive metadata of the experimental design for the accurate evaluation of the significance of the observed changes in metabolites levels are critically discussed.

Abstract 913: Differences in basal isoform phosphorylation of signaling proteins between normal and tumor tissue in different cancer types detected by the NanoPro™1000 technology

Proceedings: AACR Annual Meeting 2014; April 5-9, 2014; San Diego, CA In the era of targeted anticancer therapy, the proteomic analysis of cellular signaling networks has strongly gained in importance in the field of cancer research and treatment. There is scientific evidence that individual phosphorylation profiles of signaling proteins, besides the mutation status of patients` tumors, contain useful therapeutic information and their measurement has to be established in the clinical setting. Although, predictive genetic markers of response to targeted therapy such as KRAS testing are already in use, heterogeneity in responses to such treatment has been clinically observed. In order to understand this heterogeneity on the proteome level, the activation status of cell signaling proteins has to be analyzed. In addition, the comparison of basal phosphorylation levels of normal tissues and corresponding tumor tissues might promote a deeper understanding of cell signaling in complex clinical biospecimen. In this study, we differentially analyzed basal isoform phosphorylation in normal and tumor tissue of colorectal cancer (n= 30) and corresponding liver metastases (n= 30), as well as in breast cancer (n=26), and non-small cell lung cancer (n=29). Isoform and corresponding phosphorylation of selected key proteins from two EGF-receptor downstream pathways, i.e. Erk1/2, Mek1/2 and Akt were detected using the NanoPro™1000 technology. It has been postulated that oncogene addiction, i.e. hyper activation of oncogenic signaling pathways is a hallmark of cancer, therefore potentially restricted to tumor tissue. Today, little is known about the basal phosphorylation status of normal tissue and therapeutic consequences. In this study, we have shown that there are significant differences in the overall degree of phosphorylation between normal and tumor tissues. For example, normal liver tissue shows significant higher phosphorylation of Erk1/2 compared to liver metastases of colorectal cancer. Statistical analysis of data also revealed that the overall phosphorylation of Erk1/2, Mek1/2 and Akt significantly differed among individual patients, but not between tumor entities. Therefore, the in-depth analysis of specific isoform phosphorylation can reveal significant biological differences. Furthermore, the analysis might result in therapy relevant, predictive signatures independent of the mutation status. In summary, we found that high phosphorylation of signaling molecules is not restricted to tumor tissue and has to be taken in account when analyzing clinical biospecimen. The NanoPro TM1000 technology can be used to robustly screen a large cohort of patients, i.e. biopsies, to gain insights in phosphorylation patterns of normal and cancer tissue. Therein, the analysis of isoform phosphorylation potentially enables the selection of patients, who will most likely benefit from targeted treatment. Citation Format: Florian T. Unger, Jana Krueger, Janina Schaller, Rebecca Giese, Cordula Dede, Alexandra Samsen, Hartmut Juhl, Kerstin A. David. Differences in basal isoform phosphorylation of signaling proteins between normal and tumor tissue in different cancer types detected by the NanoPro™1000 technology. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 913. doi:10.1158/1538-7445.AM2014-913

Abstract A40: Detection of Erk1/2, Mek1/2, and Akt isoform phosphorylation in tumors of 115 patients with different cancer types using the NanoPro™1000 technology.

Proteomic analysis of cellular signaling networks has strongly gained in importance in the field of cancer research and treatment. In particular, there is a focus on the activation status of growth factor receptor related signaling cascades represented by phosphoproteins as key mediators of cellular signaling. For example, the inhibition of the EGF-receptor pathway has been widely integrated into clinical practice for different tumor entities, e.g. non small lung cancer and colorectal cancer. Although, predictive genetic markers of response to targeted therapy such as KRAS testing are already in use, heterogeneity in responses to such treatment has been clinically observed. In order to understand this heterogeneity on the proteome level, the activation status of cell signaling proteins has to be analyzed. In this study, we differentially analyzed basal isoform phosphorylation of selected key proteins from two EGF-receptor downstream pathways in breast cancer (n=26), non-small cell lung cancer (n=29), colorectal cancer (n= 30) and liver metastases of colorectal cancer (n= 30) using the NanoPro1000™ technology. Patient9s samples were collected according to Indivumed9s standard operating procedures, ensuring a high quality of tissues characterized by ischemia times below 10 minutes, comprehensive clinical data and corresponding sets of body fluids. The detailed characterization of isoform phosphorylation was conducted using established NanoPro1000™ protocols for Erk1/2, Mek1/2 and Akt. Statistical analysis of data revealed that isoform phosphorylation of Erk1/2, Mek1/2 and Akt significantly differed among individual patients, but not between tumor entities. Differences in phosphorylation patterns were characterized by the degree of overall phosphorylation, as well as individual patterns of isoform phosphorylation. This indicates that the in-depth analysis of isoform phosphorylation can reveal significant biological differences and might result in therapy relevant, predictive signatures independent of the mutation status. Hence, we established cut-off values for basal signaling molecule phosphorylation, to potentially enable the selection of patients, who will most likely benefit from targeted treatment. Based on the cut-off values we classified patients into statistical significant groups of high, moderate and low signaling molecule phosphorylation. In next steps, this classification will be further validated in terms of clinical relevance to response to targeted therapy. In summary, the NanoPro1000™ technology can be used to robustly screen a large cohort of patients for signaling molecule phosphorylation. The analysis of patients on the proteome level may be helpful to identify predictive biomarker for the development of resistance to anti-EGF-receptor treatment and thus further improve personalized medicine. Citation Information: Mol Cancer Ther 2013;12(11 Suppl):A40. Citation Format: Florian T. Unger, Jana Kruger, Janina Schaller, Giese Rebecca, Cordula Dede, Alexandra Samsen, Hartmut Juhl, Kerstin A. David. Detection of Erk1/2, Mek1/2, and Akt isoform phosphorylation in tumors of 115 patients with different cancer types using the NanoPro™1000 technology. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2013 Oct 19-23; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2013;12(11 Suppl):Abstract nr A40.

Abstract 2408: Comparison of ERK, AKT and MEK isoform phosphorylations between primary and corresponding metastatic colorectal carcinoma lesions using the NanoPro1000 technology.

Proceedings: AACR 104th Annual Meeting 2013; Apr 6-10, 2013; Washington, DC Proteomic analysis of cellular signaling networks has strongly gained in importance in the field of cancer research and treatment. In particular, there is a focus on the activation status of growth factor receptor related signaling cascades represented by phosphoproteins, as key mediators of cellular signaling. For example, the inhibition of the EGF-receptor pathway has been widely integrated into clinical practice for different tumor entities, e.g. non small lung cancer and colorectal cancer. Although, predictive genetic markers of response to targeted therapy such as KRAS testing are already in use, heterogeneity in responses to treatment has been clinically observed. In order to understand this heterogeneity on the proteome level, the activation status of cell signaling proteins has to be analyzed. The in-depth analysis of isoform phosphorylation can potentially reveal significant biological differences in activation patterns and might result in predictive signatures. In this study, we differentially analyzed basal isoform phosphorylations of selected key proteins of two EGF-receptor downstream pathways, in four primary colorectal tumors and their corresponding metastases using the NanoPro1000 technology. Patients biospecimen were collected according to standard operating procedures, ensuring a high quality of tissues characterized by ischemia times below 10 minutes, comprehensive clinical data and corresponding sets of body fluids. The detailed characterization of isoform phosphorylations was conducted using the recently introduced NanoPro1000 technology. This assay provides the separation of phospho protein isoforms by high-resolution isoelectric focusing (IEF) combined with a specific, antibody based detection. In addition to these data, the mutation status of selected, clinical relevant genes (KRAS, BRAF and PIK3CA) was determined in primary and metastatic tumor samples using Sanger sequencing. Among the analyzed primary tumors, there were two KRAS mutants, one BRAF mutant and one wild type tumor. Identical results were found in the matched metastasis. NanoPro1000 analysis revealed that isoform phosphorylation of ERK1/2, AKT and MEK1/2 differed among patients, as well as primary tumors and metastases. For example, ERK1 showed a tendency to be higher mono- and dual-phosphorylated in the primary tumors, whereas mono-phosphorylated ERK2 was elevated in metastases. However, a correlation to the mutation status was not found. These differences may explain the heterogeneity of individual responses to target anticancer treatment among patients, independent of their mutation status. Therefore, the NanoPro1000 technology can potentially be useful to identify predictive biomarker for the development of resistance to anti-EGF-receptor treatment and further improve personalized medicine. Citation Format: Florian T. Unger, Jana Krueger, Janina Schaller, Cordula Dede, Alexandra Samsen, Hartmut Juhl, Kerstin A. David. Comparison of ERK, AKT and MEK isoform phosphorylations between primary and corresponding metastatic colorectal carcinoma lesions using the NanoPro1000 technology. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 2408. doi:10.1158/1538-7445.AM2013-2408

Abstract 3605: Effects of antibody-mediated EGF-receptor inhibition on ERK 1/2 isoform phosphorylation in organoid cultures

Proceedings: AACR 103rd Annual Meeting 2012‐‐ Mar 31‐Apr 4, 2012; Chicago, IL Inhibition of the EGF-receptor activity has been widely integrated into clinical practice for different tumor entities, e.g. non small lung cancer and colorectal cancer. Besides the application of tyrosine kinase inhibitors, anti-EGF-receptor treatment is mainly based on the application of therapeutic antibodies. As a standard procedure, patients are tested for KRAS mutation and only those without mutation are selected for anti-EGF-receptor therapy. Although KRAS testing is used to guide colorectal cancer treatment, clear heterogeneity in responses to treatment has been observed. The molecular mechanisms responsible for this heterogenity are still not fully understood, In the present study, we used a preclinical model based on tumor tissue slices for a more detailed evaluation of antibody-mediated EGF-receptor inhibition. In previous experiments, the diffusion of antibodies into organoid cultures and the detection of their functional effects had been successfully shown. Tumor tissue slices from five colorectal cancer patients have been treated with an anti-EGF-receptor antibody for 48 hours. Using immunohistochemical staining, one patient showed a strong expression of the EGF-receptor, one was weak positive and three displayed a moderate staining. As readout of treatment effects, the expression and phosphorylation levels of selected key proteins from two EGF-receptor downstream pathways, the MAPK-pathway and Akt-pathway, were measured using the multiplex technology platform from Meso Scale Discovery. For a more detailed view, we have integrated the NanoPro technology in our pathway analysis approach. This new method enables the identification of distinct isoform phosphorylations and thus is able to display the complex regulation of phosphorylation patterns in detail. By using both methods, we observed a dose-dependent regulation of total- and phospho-ERK in the organoid cultures after anti-EGF-receptor treatment. Deeper analysis using the NanoPro technology showed that the regulation of isoform phosphorylation patterns of ERK1/2 differed among patients, representing the heterogeneity of individual responses among patients. In summary, the combination of the organoid drug testing platform and the NanoPro technology can potentially be useful to identify predictive phosphorylation patterns of tumor cells within their natural tumor microenvironment. This may facilitate the discovery of predictive biomarker for the response to anti-EGF-receptor treatment and further improve personalized medicine. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 3605. doi:1538-7445.AM2012-3605

Abstract C228: Preclinical analysis of therapeutic antibody efficiancy using an organoid model-based drug testing platform.

A common problem in translational research is the fact that findings from cell line experiments cannot be transferred one to one to the clinical setting, because each tumor consists of a variety of different cancer cells interacting with stroma cells, resulting in a highly variable response to drugs. In order to improve this situation, we previously developed an organoid culture based, preclinical drug testing platform suitable for the investigation of anticancer drug effects (cytotoxic drugs). In this study, we further advanced this model in order to measure effects of targeted drugs, such as small molecules and antibodies. Here we show that effects of therapeutic antibodies such as Trastuzumab (Herceptin®) and small molecules such as Gefitinib (Iressa®) were detectable using this model. Tumor tissue slices from colorectal and breast cancer patients were prepared and cultured according to Indivumed9s standard operating protocols. First, we determined the degree of antibody diffusion into organoid cultures. Tissue slices were incubated with a commercially available antibody followed by immunofluorescence staining and microscopy. Functional effects of anti-HER-2 (Trastuzumab) treatment were examined in HER-2 positive and negative breast cancer tissue slices. In parallel, results have been confirmed in breast cancer cell lines (BT474 and MCF-7 cells) with positive and negative HER-2 expression. Trastuzumab treatment of breast cancer tissue slices and breast cancer cell lines showed a dose-dependent reduction of pAkt only in HER-2 positive tissues and cells. HER-2 expression levels of tissue slices and cells were previously analyzed by immunohistochemistry (IHC). IHC is a special feature of this model due to the capability to visualize drug effects within a tumor in individual cells. In summary, the data revealed that large molecules such as antibodies diffuse into 400 μm thick cultured tissue slices reaching their target within 24 hours. Furthermore, we demonstrated that functional drug effects of therapeutic antibodies measured in secondary cell lines were reproduced in organoid cultures. The preclinical drug testing platform developed by Indivumed enables the examination of effects of classical chemotherapeutics and small molecules, as we have shown recently, as well as effects of larger molecules, such as antibodies. This implies that the model is suitable for a more comprehensive analysis of drug responses and in particular cellular responses to targeted drugs in a natural tumor microenvironment. This will ultimately allow for the determination of efficiency of anticancer drugs, help to prioritise drug candidates, support dose finding, identify promising indications for clinical trails and guide individualized therapy. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2011 Nov 12-16; San Francisco, CA. Philadelphia (PA): AACR; Mol Cancer Ther 2011;10(11 Suppl):Abstract nr C228.

Abstract 3177: Early detection biomarker for colorectal malignancies identified by a novel comparative proteomic workflow

Colorectal cancer (CRC) ranks worldwide at third place for tumor diseases and account for an annual mortality rate of 492.000 cases. Despite the advances in treatment of CRC, patients mostly benefit from an early detection of malignancy. Although several molecular events are known to be associated with tumor development and progression the accuracy of currently practiced early detection of CRC has to be optimized. Therefore, the discovery of new and more reliable biomarker for diagnosis of early stages of cancer is still a challenge. Towards this end, our study was designed to identify protein biomarkers which are closely associated with the molecular events leading from epithelial adenoma to the early stages of carcinoma. Tissues from patients with epithelial adenoma (high dysplasia), carcinoma (stage I) and matched normal tissue (n = 10 for each group) were selected and analysed using our novel biomarker discovery platform. The workflow combines the cell specificity and the selectivity of laser capture microdissection (LCM) with the resolution power and sensitivity of liquid-chromatography (LC)-matrix-assisted-laser-desorption/ionization mass spectrometry (LC-MALDI-MS). According to this LCM-LC-MALDI-MS strategy microdissected cells were lysed and extracted proteins were digested with trypsin. Obtained peptides were separated by capillary reversed phase HPLC. Tryptic fragments were subsequently detected by reflector MALDI-MS measurements. Differential peptide analysis was performed to discover robust and significant expression differences between patient groups. Selected peptides were subsequently fragmented by MS/MS experiments to reveal protein identity. Up to 7000 ion signals ranging from m/z 800 to 4000 per sample were generated and used for statistical analysis. Our targeted biomarker discovery approach resulted in the identification of more than 30 biomarker candidates. The identified early detection biomarker candidates are involved in diverse cellular functions and can be linked to early or late events of tumor progression. Among these hNRPU, HSP90 and SFPQ were so far further analysed. All three biomarker candidates were significantly regulated between groups, showing increasing expression from normal across adenoma to carcinoma tissue. Currently these biomarkers are being validated using antibody based assays. In this study we were able to show that the newly established high resolution proteomic workflow is applicable for the detection and identification of regulated peptides in microdissected tissue compartments. In summary the established comparative proteomic workflow is suitable for the discovery of biomarker for miscellaneous applications (e.g. early detection, prognosis and prediction of response to therapy) Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 102nd Annual Meeting of the American Association for Cancer Research; 2011 Apr 2-6; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2011;71(8 Suppl):Abstract nr 3177. doi:10.1158/1538-7445.AM2011-3177

Abstract B43: Comparative proteomic analysis of colorectal cell lines to identify biomarkers for the prediction of intrinsic resistance to FOLFOX chemotherapy

As there is a great interpatient variability in drug response caused by the individual geno- and phenotype the choice of optimal chemotherapy should be tailored for each patient. Intrinsic drug resistance is one of the main reasons for therapeutic ineffectiveness of first line therapy and recurrence in colorectal cancer. To advance individualized chemotherapy and understanding of drug resistance mechanisms, predictive biomarkers which indicate individual drug response of patients are urgently needed. The prediction of response to first line therapy will identify subgroups of patients who benefit from treatment and thus avoid ineffective treatment and associated side effects. In this study, several commercially available cell lines as well as primary epithelial mixed cultures and clonal cell lines established from colorectal cancer patients were treated with different drugs, such as 5-fluorouracil, oxaliplatin, and drug combinations, e.g., FOLFOX. The determination of chemosensitivity of these cell lines to drug treatment was performed using two different assays, the ATPlite™-(Perkin Elmer) and FMC-assay (according to Larsson et al.). Both assays showed time- and dose-dependent drug responses which differed significantly among the 18 colorectal cell cultures. The analysis of obtained in vitro chemosensitivity data allowed us to cluster sensitive versus resistant cell lines by determining IC50-values. For the discovery of differentially expressed proteins which might be linked to a specific phenotypes of intrinsic chemoresistance we generated protein expression patterns from all 18 cell lines using a combination of reversed phase highperformance liquid chromatography and matrix-assisted laser desorption/ionization mass spectrometry (RP-HPLC MALDI MS). Up to 3500 m/z ion signals in the molecular mass range between 2500-35000 m/z were detected reflecting a complex pattern of small proteins and peptides. For protein identification, RP-HPLC fractionated samples were purified and digested by trypsin. Fragments of interest were subjected to MS/MS analysis in the TOF/TOF mode. The corresponding MS/MS spectra were used to search the NCBInr database using MASCOT software. Based on these data, we hope to gain more understanding of the biological processes of chemoresistance concerning important cellular drug responses (e.g., signal transduction pathways) and thereby obtain insight into new drug targets for future therapeutic intervention. Citation Information: Clin Cancer Res 2010;16(14 Suppl):B43.

Abstract 4577: Comparative proteomic biomarker screening in laser microdissected tissues from colon adenoma and cancer

Proceedings: AACR 101st Annual Meeting 2010‐‐ Apr 17‐21, 2010; Washington, DC Colorectal (CRC) malignancies rank world wide at the third place for tumor diseases and account for an annual mortality rate of 492 000 cases. Although several molecular events are known to be involved in the transition from normal tissue to adenoma and finally to undifferentiated carcinoma it remains a challenge to discover new and more reliable biomarkers for diagnosis, prognosis and prediction of outcome. Towards this end a study was designed to identify potential biomarkers which are associated with the molecular events leading from epithelial adenoma to the early stages of carcinoma. A new biomarker discovery strategy was developed to combine the cell specificity and selectivity of laser capture microdissection (LCM) with the resolution power and sensitivity of liquid-chromatography (LC)-matrix-assisted-laser-desorption/ionization mass spectrometry (LC-MALDI-MS). We carefully selected a group of closely matched patients (n = 10 for each group) afflicted with epithelial adenoma (high dysplasia) or early stages of carcinoma (stage I) and used the derived normal as well as the matched tumor tissue samples to reveal protein expression differences. According to this LC-MALDI-MS strategy microdissected cells (around 10 000 cells) were lysed and extracted proteins were digested with Trypsin. Obtained peptides were separated by capillary reversed phase HPLC (Agilent). The resulting LC-fractions (300) were spotted on prespotted AnchorChip targets (PAC, Bruker) and tryptic fragments subsequently detected by reflector MALDI-MS (ultraflex III, Bruker) measurements. This generated between 5000-7000 ion signals ranging from m/z 800 to 4000. Differential peptide analysis was then performed with the goal to discover robust and significant expression differences between patient groups. Therefore, only m/z ions displaying a minimum twofold difference and a p-value of 0.01 between the tissue samples were considered for further analysis. The selected peptides were subsequently fragmented by MS/MS experiments to reveal their primary sequence and protein identity. Our targeted biomarker discovery approach resulted in the identification of more than 30 biomarker candidates which act in diverse cellular functions and can now be linked to early events (e.g. adenoma vs. normal tissue) or later events (e.g. carcinoma vs. normal) of tumor progression. Currently these biomarkers are validated using antibody based assays to further analyze their potential as markers in a clinical setting. Note: This abstract was not presented at the AACR 101st Annual Meeting 2010 because the presenter was unable to attend. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 4577.