Supplementary Figure 1 from Protein Expression Profiling in High-Risk Breast Cancer Patients Treated with High-Dose or Conventional Dose–Dense Chemotherapy

Supplementary Table 1 from Expression Profiling of t(12;22) Positive Clear Cell Sarcoma of Soft Tissue Cell Lines Reveals Characteristic Up-Regulation of Potential New Marker Genes Including <b><i>ERBB3</i></b>

Driven by genomic somatic variation, tumour tissues are typically heterogeneous, yet unbiased quantitative methods are rarely used to analyse heterogeneity at the protein level. Motivated by this problem, we developed automated image segmentation of images of multiple biomarkers in Ewing sarcoma to generate distributions of biomarkers between and within tumour cells. We further integrate high dimensional data with patient clinical outcomes utilising random survival forest (RSF) machine learning. Using material from cohorts of genetically diagnosed Ewing sarcoma with EWSR1 chromosomal translocations, confocal images of tissue microarrays were segmented with level sets and watershed algorithms. Each cell nucleus and cytoplasm were identified in relation to DAPI and CD99, respectively, and protein biomarkers (e.g. Ki67, pS6, Foxo3a, EGR1, MAPK) localised relative to nuclear and cytoplasmic regions of each cell in order to generate image feature distributions. The image distribution features were analysed with RSF in relation to known overall patient survival from three separate cohorts (185 informative cases). Variation in pre-analytical processing resulted in elimination of a high number of non-informative images that had poor DAPI localisation or biomarker preservation (67 cases, 36%). The distribution of image features for biomarkers in the remaining high quality material (118 cases, 104 features per case) were analysed by RSF with feature selection, and performance assessed using internal cross-validation, rather than a separate validation cohort. A prognostic classifier for Ewing sarcoma with low cross-validation error rates (0.36) was comprised of multiple features, including the Ki67 proliferative marker and a sub-population of cells with low cytoplasmic/nuclear ratio of CD99. Through elimination of bias, the evaluation of high-dimensionality biomarker distribution within cell populations of a tumour using random forest analysis in quality controlled tumour material could be achieved. Such an automated and integrated methodology has potential application in the identification of prognostic classifiers based on tumour cell heterogeneity.

Introduction: Head and neck squamous cell carcinomas (HNSSCs) are one of the leading causes of cancer-associated death worldwide. Although certain behavioral risk factors are well recognized as tumor promoting, there is very little known about the presence of predisposing germline mutations in HNSCC patients.Methods: In this study, we analyzed 121 individuals with HNSCCs collected at our institution for germline alterations in the newly identified cancer susceptibility gene RAD51C.Results: Sequencing of all exons and the adjacent introns revealed five distinct heterozygous sequence deviations in RAD51C in seven patients (5.8%). A female patient without any other risk factors carried a germline mutation that disrupted the canonical splice acceptor site of exon 5 (c.706-2A>G).Conclusions: As there are only a few publications in the literature identifying germline mutations in head and neck cancer patients, our results provide the first indication that paralogs of RAD51, recently described as mutated in breast and ovarian cancer patients, might also be candidates for genetic risk factors in sporadic squamous cell carcinomas of the head and neck. (C) 2013 Elsevier Ltd. All rights reserved.

Three-dimensional (3D) tumor cell cultures grown in laminin-rich-extracellular matrix (lrECM) are considered to reflect human tumors more realistic as compared to cells grown as monolayer on plastic. Here, we systematically investigated the impact of ECM on phenotype, gene expression, EGFR signaling pathway, and on EGFR inhibition in commonly used colorectal cancer (CRC) cell lines. LrECM on-top (3D) culture assays were performed with the CRC cell lines SW-480, HT-29, DLD-1, LOVO, CACO-2, COLO-205 and COLO-206F. Morphology of lrECM cultivated CRC cell lines was determined by phase contrast and confocal laser scanning fluorescence microscopy. Proliferation of cells was examined by MTT assay, invasive capacity of the cell lines was assayed using Matrigel-coated Boyden chambers, and migratory activity was determined employing the Fence assay. Differential gene expression was analyzed at the transcriptional level by the Agilent array platform. EGFR was inhibited by using the specific small molecule inhibitor AG1478. A specific spheroid growth pattern was observed for all investigated CRC cell lines. DLD-1, HT-29 and SW-480 and CACO-2 exhibited a clear solid tumor cell formation, while LOVO, COLO-205 and COLO-206F were characterized by forming grape-like structures. Although the occurrence of a spheroid morphology did not correlate with an altered migratory, invasive, or proliferative capacity of CRC cell lines, gene expression was clearly altered in cells grown on lrECM as compared to 2D cultures. Interestingly, in KRAS wild-type cell lines, inhibition of EGFR was less effective in lrECM (3D) cultures as compared to 2D cell cultures. Thus, comparing both 2D and 3D cell culture models, our data support the influence of the ECM on cancer growth. Compared to conventional 2D cell culture, the lrECM (3D) cell culture model offers the opportunity to investigate permanent CRC cell lines under more physiological conditions, i.e. in the context of molecular therapeutic targets and their pharmacological inhibition.

Purpose Targeted therapies using the anti-EGFR antibodies panitumumab (Pmab) or cetuximab (Cmab) are currently restricted to patients with metastatic colorectal adenocarcinoma whose tumours do not show a mutation in KRAS. However, recent retrospective studies indicated that patients with tumours mutated in codon 13 of KRAS may benefit from treatment with Cmab in contrast to patients with tumours mutated in KRAS codon 12. Methods To study the functional impact of the subtype of KRAS mutations on the efficiency of EGFR-targeted therapies, we correlated the KRAS mutation status of 15 colorectal carcinoma cell lines with the in vitro sensitivity of these cells to Cmab/Pmab. Mutations in the potential predictive biomarkers BRAF and PIK3CA as well as protein expression of EGFR and PTEN were also determined. Results Four out of seven KRAS-mutated cell lines were characterised by the p.G13D mutation. Treatment of these cells using Cmab/Pmab induced a significant growth inhibition in contrast to cell lines showing a KRAS mutation at codon 12 or 61. Out of the eight KRAS wild-type cell lines, five were insensitive to Cmab/Pmab. These cell lines were characterised either by BRAF mutation or by absence of EGFR or PTEN protein expression. Conclusions Since KRAS p.G13D-mutated tumour cells may respond to EGFR-targeted therapy, we suggest including subtype analysis of KRAS mutations in prospective clinical trials. In KRAS wild-type tumour cells, BRAF mutations and loss of EGFR or PTEN expression may lead to resistance to EGFR-targeted therapy and should be considered as additional negative predictive biomarkers.

Background Chemokine receptor CXCR4, together with its ligand CXCL12, plays critical roles in cancer progression, including growth, metastasis and angiogenesis. Ewing sarcoma is a sarcoma with poor prognosis despite current therapies, particularly for patients with advanced-stage disease. Lungs and bone (marrow), organs of predilection for (primary/metastatic) Ewing sarcoma, represent predominant CXCL12 sources. Methods To gain insight into the role of the CXCR4-CXCL12 axis in Ewing sarcoma, CXCR4, CXCL12 and hypoxia-inducible factor-1α protein expression was studied in therapy-naïve and metastatic tumors by immunohistochemistry. CXCR4 function was assessed in vitro , by flow cytometry and proliferation/ cell viability assays, in the presence of recombinant CXCL12 and/or CXCR4-antagonist AMD3100 or under hypoxic conditions. Results Whereas CXCR4 was predominantly expressed by tumor cells, CXCL12 was observed in both tumor and stromal areas. Survival analysis revealed an (expression level-dependent) negative impact of CXCR4 expression (p < 0.04). A role for the CXCR4-CXCL12 axis in Ewing sarcoma growth was suggested by our observations that i) CXCR4 expression correlated positively with tumor volume at diagnosis (p = 0.013), ii) CXCL12 was present within the microenvironment of virtually all cases, iii) CXCL12 induced proliferation of CXCR4-positive Ewing sarcoma cell lines, which could be abrogated by AMD3100. CXCR4 expression was not correlated with occurrence of metastatic disease. Also, therapy-naïve tumors demonstrated higher CXCR4 expression as compared to metastases (p = 0.027). Evaluation of in vivo hypoxia-inducible factor-1α expression and culture of cells under hypoxic conditions revealed no role for hypoxia in CXCR4 expression. Conclusions Together, our results imply a crucial role for the CXCR4-CXCL12 axis in auto- and/or paracrine growth stimulation. Integration of CXCR4-targeting strategies into first- and/or second-line treatment regimens may represent a promising treatment option for Ewing sarcoma.

Identification of factors to detect chemotherapy-resistant tumours at diagnosis is a first priority for risk-adapted therapy in the oncology of children and young adults, where more individualized, effective, and less toxic treatments are highly desirable. In this study, we analysed the miRNAs discriminating Ewing's sarcoma (EWS) patients with different clinical outcomes in order to identify new indicators of prognosis. miRNA expression was investigated in 49 primary EWSs by using the Agilent human miRNA microarray v.2 and/or qRT-PCR. Statistical power of the samples studied for miRNA expression was verified, indicating adequate sample size. Microarray analysis defined a signature of five miRNAs (miR-34a, miR-23a, miR-92a, miR-490-3p, and miR-130b) as an independent predictor of risk for disease progression and survival. Validation analysis in the extended sample set indicated that both miR-34a and miR-490-3p achieved sufficient statistical power to predict prognosis. Results were particularly robust for miR-34a, which appeared associated with either event-free or overall survival and emerged as a significant predictor also after multivariate analysis. Patients with the highest expression of miR-34a did not experience adverse events in 5 years; in contrast, patients with the lowest expression recurred within 2 years. High expression of miR34a can be detected also in paraffin-embedded tissues by in situ hybridization, thus contributing to an easy routine evaluation of this miRNA. Functional analysis of miR-34a in EWS cell lines indicated that when miR-34a expression was enforced, cells were less proliferative, less malignant, and sensitized to doxorubicin and vincristine. Expression of miR-34a could be increased in p53wt cells by treatment with nutlin-3a. Accordingly, nutlin-3a synergizes with doxorubicin. Overall, our data indicate that miR-34a expression is a strong predictor of outcome in EWS. Restoration of miR-34a activity may be useful to decrease malignancy and increase tumour sensitivity to current drugs, so sparing excessive long-term toxicity to EWS patients.

Cancer cell lines represent in vitro models for studying malignancies, general cell biology, drug discovery and more. Whether they can be considered as exact representative models of the parental tumors remains uncertain given the acquisition of additional ex vivo changes of the cells and the lack of tissue architecture and stroma. Previously, within the EuroBoNeT consortium, we characterized a collection of bone sarcoma cell lines on genomic and proteomic level. Here, we address the phenotypical and functional characterization of the unique set of osteosarcoma cell lines (n=19) in vitro and in vivo. For functional analysis of differentiation capacity, cells were stimulated towards osteoblasts, adipocytes and chondrocytes. Furthermore, all cell lines were injected subcutaneously and intramuscularly into nude mice to assay their in vivo tumor formation capacity as well as for phenotypical analysis of the tumors. All formed tumors were further characterized histologically and immunohistochemically. Out of 19 cell lines, 17 (89%) showed adipogenic differentiation, 13/19 (68%) could differentiate towards osteoblasts and in 6/19 (32%) cell lines chondrogenic differentiation was evident. About half of the cell lines (8/19, 42%) produced tumors in vivo after subcutaneous and intramuscular injections. Several cell lines showed invasion into adjacent tissues and one tumor developed several lung metastases. The use of cell lines, especially in cancer research, is of paramount importance. Here, we identify comprehensively characterized osteosarcoma cell lines, which robustly represent clinical osteosarcoma providing researchers useful in vitro and in vivo models to study the genetics and functional characteristics of this highly malignant neoplasm.

Alveolar Rhabdomyosarcomas (RMA) are characterized by chromosomal translocations, fusing the PAX3 or PAX7 gene with FKHR in about 85%. Previous studies have suggested that the fusion type is associated with prognosis. In order to investigate the predictive value of the PAX–FKHR fusion status on disease outcome of patients with RMA treated in the CWS trials we performed a retrospective analysis.

PURPOSE EWS-ETS fusion genes are the driving force in Ewing's sarcoma pathogenesis. Because of the variable breakpoint locations in the involved genes, there is heterogeneity in fusion RNA and protein architecture. Since previous retrospective studies suggested prognostic differences among patients expressing different EWS-FLI1 fusion types, the impact of fusion RNA architecture on disease progression and relapse was studied prospectively within the Euro-E.W.I.N.G. 99 clinical trial. PATIENTS AND METHODS Among 1,957 patients who registered before January 1, 2007, 703 primary tumors were accessible for the molecular biology study. Fusion type was assessed by polymerase chain reaction on frozen (n = 578) or paraffin-embedded materials (n = 125). The primary end point was the time to disease progression or relapse. Results After exclusion of noninformative patients, 565 patients were entered into the prognostic factor analysis comparing type 1 (n = 296), type 2 (n = 133), nontype 1/nontype 2 EWS-FLI1 (n = 91) and EWS-ERG fusions (n = 45). Median follow-up time was 4.5 years. The distribution of sex, age, tumor volume, tumor site, disease extension, or histologic response did not differ between the four fusion type groups. We did not observe any significant prognostic value of the fusion type on the risk of progression or relapse. The only slight difference was that the risk of progression or relapse associated with nontype 1/nontype 2 EWS-FLI1 fusions was 1.38 (95% CI, 0.96 to 2.0) times higher than risk associated with other fusion types, but it was not significant (P = .10). CONCLUSION In contrast to retrospective studies, the prospective evaluation did not confirm a prognostic benefit for type 1 EWS-FLI1 fusions.

Purpose: Epidermal growth factor receptor (EGFR) antibody therapy is established in patients with wild-type KRAS colorectal carcinoma; however, up to 50% of these patients do not respond to this therapy. To identify the possible causes of this therapy failure, we searched for mutations in different EGFR-dependent signaling proteins and analyzed their distribution patterns in primary tumors and corresponding metastases.Experimental Design: Tumor tissues, macrodissected from tumor centers, invasion fronts (n = 100), lymph nodes (n = 55), and distant metastases (n = 20), respectively, were subjected to DNA extraction and mutation analysis of KRAS, BRAF, and PIK3CA.Results: Activating mutations were detected in 41% (KRAS), 7% (BRAF), and 21% (PIK3CA) of the primary tumors. By comparing tumor centers and invasion fronts, the intratumoral heterogeneity of KRAS, BRAF, and PIK3CA mutations was observed in 8%, 1%, and 5% of primary tumors, respectively. Heterogeneity between primary tumors and lymph node metastases was found in 31% (KRAS), 4% (BRAF), and 13% (PIK3CA) of the cases. Heterogeneity between primary tumors and distant metastases was present in two patients (10%) for KRAS and one patient for PIK3CA (5%), but not for BRAF. Discordant results between primary tumors and metastases could markedly be reduced by testing the additional tumor samples.Conclusions: Failure of EGFR antibody therapy in patients with wild-type KRAS colorectal cancer may result from activating BRAF or PIK3CA mutations and false-negative sequencing results caused by intratumoral heterogeneity. Due to the particularly high rates of heterogeneity between primary tumors and lymph node metastases, the latter are least suitable for diagnostic mutation analysis. Clin Cancer Res; 16(3); 790-9. (C) 2010 AACR.

Hypoxia regulates gene expression via the transcription factor HIF (Hypoxia-Inducible Factor). Little is known regarding HIF expression and function in primary bone sarcomas. We describe HIF expression and phenotypic effects of hypoxia, hypoglycaemia and HIF in Ewing's sarcoma and osteosarcoma.

Usage of cancer cell lines has repeatedly generated conflicting results provoked by differences among subclones or contamination with mycoplasm or other immortal mammalian cells. To overcome these limitations, we decided within the EuroBoNeT consortium to characterize a common set of cell lines including osteosarcomas (OS), Ewing sarcomas (ES), and chondrosarcomas (CS). DNA fingerprinting was used to guarantee the identity of all of the cell lines and to distinguish subclones of osteosarcoma cell line HOS. Screening for homozygous loss of 38 tumor suppressor genes by MLPA revealed deletion of CDKN2A as the most common event (15/36), strictly associated with absence of the CDKN2A (p16) protein. Ten cell lines showed missense mutations of the TP53 gene while another set of nine cell lines showed mutations resulting in truncation of the TP53 protein. Cells harboring missense mutations expressed high levels of nuclear TP53, while cell lines with nonsense mutations showed weak/absent staining for TP53. TP53wt cell lines usually expressed the protein in 2–10% of the cells. However, seven TP53wt osteosarcomas were negative for both mRNA and protein expression. Our analyses shed light on the correlation between immunohistochemical and genetic data for CDKN2A and TP53, and confirm the importance of these signaling pathways. The characterization of a substantial number of cell lines represents an important step to supply research groups with proven models for further advanced studies on tumor biology and may help to make results from different laboratories more comparable. © 2009 Wiley‐Liss, Inc.

Ewing sarcoma (EWS) is a tumour most commonly arising in bone, although on occasion in soft tissue, with a poor prognosis in patients with refractory or relapsed disease, despite multimodal therapy. Immunotherapeutic strategies based on tumour-reactive T and/or natural killer cells may improve the treatment of advanced-stage EWS. Since cellular immune recognition critically depends on human leukocyte antigen (HLA) expression, knowledge about HLA expression in EWS is crucial in the design of cellular immunotherapeutic strategies. Constitutive and IFNgamma-induced HLA class I expression was analysed in EWS cell lines (n = 6) by flow cytometry, using antibodies against both monomorphic and allele-specific antigens. Expression of antigen processing pathway components and beta-2 microglobulin (beta2m) was assessed by western blot. Expression of class II transactivator (CIITA), and its contribution to HLA class II expression, was evaluated by qRT-PCR, transduction assays, and flow cytometry. beta2m/HLA class I and class II expression was validated in EWS tumours (n = 67) by immunofluorescence. Complete or partial absence of HLA class I expression was observed in 79% of EWS tumours. Lung metastases consistently lacked HLA class I and sequential tumours demonstrated a tendency towards decreased expression upon disease progression. Together with absent or low constitutive expression levels of specific HLA class I loci and alleles, and differential induction of identical alleles by IFNgamma in different cell lines, these results may reflect the existence of an immune escape mechanism. Inducible expression of TAP-1/-2, tapasin, LMP-2/-7, and the beta2m/HLA class I complex by IFNgamma suggests that regulatory mechanisms are mainly responsible for heterogeneity in constitutive class I expression. EWSs lack IFNgamma-inducible HLA class II, due to lack of functional CIITA. The majority of EWS tumours, particularly if advanced-stage, exhibit complete or partial absence of both classes of HLA. This knowledge will be instrumental in the design of cellular immunotherapeutic strategies for advanced-stage EWS.

BACKGROUNDFormalin-fixed paraffin-embedded (FFPE) tumor material represents a valuable resource for the analysis of RNA-based biomarkers, both in research laboratories and in routine clinical testing. A robust and automated RNA-extraction method with a high sample throughput is required.METHODSWe evaluated extraction performance for 4 silica-based RNA-extraction protocols: (a) a fully automated, bead-based RNA-isolation procedure; (b) its manual counterpart; (c) a semiautomated bead-based extraction system; and (d) a manual column-based extraction kit. RNA from 360 sections (90 sections per extraction method) of 30 FFPE tumor blocks up to 20 years of age was purified and analyzed by quantitative reverse-transcription PCR for ESR1 (estrogen receptor 1), PGR (progesterone receptor), ERBB2 [v-erb-b2 erythroblastic leukemia viral oncogene homolog 2, neuro/glioblastoma derived oncogene homolog (avian)], and RPL37A (ribosomal protein L37a).RESULTSThe semiautomated protocol gave the best yield. The 3 bead-based methods showed good across-method correlations in both yield and relative mRNA amounts (r = 0.86-0.95 and 0.98, respectively). In contrast, correlations between any of the bead-based methods and the manual column-based method were worse (r = 0.77-0.95 and 0.96, respectively). The fully automated method showed the lowest variation from section to section (root mean square error, 0.32-0.35 Cq, where Cq is the quantification cycle) and required the least hands-on time (1 h).CONCLUSIONSThe fully automated RNA-purification method showed the best reproducibility in gene expression analyses of neighboring sections of tissue blocks between 3 and 20 years of age and required the least overall and hands-on times. This method appears well suited for high-throughput RNA analyses in both routine clinical testing and translational research studies with archived FFPE material.