Identification of T cell target antigens in glioblastoma stem-like cells using an integrated proteomics-based approach in patient specimens.

Glioblastoma (GBM) is a highly aggressive brain tumor and still remains incurable. Among others, an immature subpopulation of self-renewing and therapy-resistant tumor cells-often referred to as glioblastoma stem-like cells (GSCs)-has been shown to contribute to disease recurrence. To target these cells personalized immunotherapy has gained a lot of interest, e.g. by reactivating pre-existing anti-tumor immune responses against GSC antigens. To identify T cell targets commonly presented by GSCs and their differentiated counterpart, we used a proteomics-based separation of GSC proteins in combination with a T cell activation assay. Altogether, 713 proteins were identified by LC-ESI-MS/MS mass spectrometry. After a thorough filtering process, 32 proteins were chosen for further analyses. Immunogenicity of corresponding peptides was tested ex vivo. A considerable number of these antigens induced T cell responses in GBM patients but not in healthy donors. Moreover, most of them were overexpressed in primary GBM and also highly expressed in recurrent GBM tissues. Interestingly, expression of the most frequent T cell target antigens could also be confirmed in quiescent, slow-cycling GSCs isolated in high purity by the DEPArray technology. Finally, for a subset of these T cell target antigens, an association between expression levels and higher T cell infiltration as well as an increased expression of positive immune modulators was observed. In summary, we identified novel immunogenic proteins, which frequently induce tumor-specific T cell responses in GBM patients and were also detected in vitro in therapy-resistant quiescent, slow-cycling GSCs. Stable expression of these T cell targets in primary and recurrent GBM support their suitability for future clinical use.

Expression and regulation of AC133 and CD133 in glioblastoma.

The biological significance of CD133 in glioblastoma is controversial. Above all, there is disagreement concerning the proper approach, the appropriate (cell) model and the suitable microenvironment to study this molecule, often leading to inconsistent experimental results among studies. In consideration of a primary need to dissect and to understand the CD133 phenotype in glioblastoma we performed a comprehensive analysis of CD133 expression and regulation in a large set of glioblastoma cell lines (n = 20) as well as in tumor xenografts. Our analysis considered alternatively spliced mRNA transcripts, different protein epitopes as well as varying sub-cellular localizations of CD133 and explored its regulation under pertinent micro-environmental conditions. CD133 mRNA and CD133 protein could be detected in all relevant types of glioblastoma cell lines. In addition, we detected frequent intracellular CD133 protein accumulations located to the ER and/or Golgi apparatus but seemingly unrelated to particular CD133 splice variants or protein epitopes. In contrast, membrane-bound expression of CD133 was restricted to tumor cells bearing the extracellular CD133 epitope AC133. Only in these cells, differentiation and oxygen levels clearly impacted on AC133 expression and to some extent also influenced CD133 mRNA and protein expression. Most importantly, however, modulation of AC133 levels could occur independently of changes in CD133 mRNA transcription, CD133 protein translation, protein retention or protein shedding. Our results suggest that the AC133 epitope, rather than CD133 mRNA or protein, mirrors malignancy-related tumor traits such as tumor differentiation and local oxygen tension levels, and thus corroborate its role as a biologically relevant cancer marker.