NIH Image to ImageJ: 25 years of image analysis. are stable, making them suitable for research purposes as well as drug development process utilizing high throughput approach. (is accompanied in 35-60% of cases by expression of an oncogenic mutant receptor, termed variant III or vIII, that is unique to tumor tissues making it an attractive therapeutic target [12, 13]. Characterized by intragenic deletion of exons 2-7, which constitute the ligand binding domain, EGFRvIII is described as constitutively active receptor. Investigation of EGFRvIII expression in tumor tissue reveals a distinct pattern, with only a small portion of cells being positive for mutant receptor expression [6, 14, 15]. The effects of aberrant signaling by EGFRvIII have been reported to be cell intrinsic as well as extrinsic, with a number of secreted growth factors and cytokines described [16-20]. Both autocrine as well as paracrine signaling are associated with EGFRvIII expression, leading to increased cancer cell growth, survival, proliferation and altered metabolism [21-23]. Also invasiveness of cancer cells expressing EGFRvIII is elevated, with positive correlation in expression of a number of metalloproteinases, MMP-9 in particular [7, 24]. Moreover, dynamic regulation of the amplicon number has been reported to mediate drug resistance of glioblastoma cells [5, 25]. Taken together, those characteristics define EGFRvIII as a potent oncogene and attractive therapeutic target. At present, no therapies targeting EGFRvIII are used in the clinic. One of the reasons for this is lack of appropriate models to study the biology of the receptor and, more importantly, develop novel therapeutics. Difficulties associated with establishment of EGFRvIII expressing GB models are related to the loss of and amplicons during the stabilization process, causes of which (E)-ZL0420 are unknown [26, 27]. For this reason, neurospheres from primary cancer cells or xenografts thereof are commonly used for research purposes . Unfortunately, low material availability, low stability of the model (neurospheres) or high associated costs (xenografts) make those models inappropriate for drug development process, especially at the early stages of development [26, 28-30]. Alternatively, stable cell lines genetically modified to express EGFRvIII are used , however, such models do not account for tumor tissue heterogeneity or extrachromosomal nature of and is suitable for high throughput studies utilized in drug development. RESULTS Analysis of currently (E)-ZL0420 used glioblastoma models Investigation of the protein activity is best conducted in the environment as close to the native as possible, allowing for insight into the functional biology of the protein. Therefore, we have attempted using neurospheres formed by primary cell cultures obtained from surgical resections. Despite problems with stabilization of the primary cell cultures reported previously , we have analyzed nine glioblastoma resections, two of which were positive for EGFRvIII transcript (Figure ?(Figure1A).1A). Treatment of EGFRvIII-positive neurospheres with erlotinib produced variable results between tumors (Figure ?(Figure1B1B and Sup.Figure 1A). Analogous situation was observed upon treatment with EGF, with 50% of spheres from the same tumor not showing any effect and the remaining ones displaying signs of cell death (Sup.Figure 1B). Our attempts at stabilization of the primary glioblastoma cells positive for EGFRvIII in the form of an adherent cell line was only partially successful for only one of the tumors, with cancer cells surviving post-passage 10 without amplicons. RT-PCR analysis of the EGFRvIII mRNA levels clearly indicated a rapid decline (Figure ?(Figure1C),1C), consistent with reports in the literature [26, 27]. Open in a separate window Figure 1 Assessment of models currently used to study EGFRvIIIA. Glioblastoma samples were analyzed on the mRNA level for EGFRvIII and EGFRWT expression. B. Neurospheres obtained from glioblastoma resections positive for EGFRvIII expression were treated with DMSO or erlotinib (10 M). At least 3 neurospheres were analysed in each condition. C. Adherent cell line established from ARAD31 was cultured over several passages and 0.05; ns, not significant. With stable cell lines offering a less variable model, we attempted inserting (E)-ZL0420 cDNA under the control of the constitutively active CMV promoter into U87-MG and NCI-H460 cell lines using lipofection or lentiviral transduction, respectively. A couple of stable Synpo clones were established from both cell lines, (E)-ZL0420 however, expression of the transgene varied among them on the mRNA level, despite the same transfection protocol (Figure ?(Figure1D).1D). Assessment of EGFRvIII expression on the protein level in H460 line using western blotting proved impossible, as a nonspecific band.