Article
EGFR-associated intratumoral heterogeneity – impact on tumor initiation, progression and response to targeted therapy deciphered by multicolor labeling of EGFR-amplified glioblastoma cells
Search Medline for
Authors
-
Katrin Liffers
- Klinik und Poliklinik für Neurochirurgie, Labor für Hirntumorbiologie, Universitätsklinikum Hamburg-Eppendorf
-
Kristoffer Riecken
- Interdisziplinäre Klinik und Poliklinik für Stammzelltransplantation, Forschungsabteilung für Zell- und Gentherapie, Universitätsklinikum Hamburg-Eppendorf
-
Boris Fehse
- Interdisziplinäre Klinik und Poliklinik für Stammzelltransplantation, Forschungsabteilung für Zell- und Gentherapie, Universitätsklinikum Hamburg-Eppendorf
-
Manfred Westphal
- Klinik und Poliklinik für Neurochirurgie, Labor für Hirntumorbiologie, Universitätsklinikum Hamburg-Eppendorf
-
Katrin Lamszus
- Klinik und Poliklinik für Neurochirurgie, Labor für Hirntumorbiologie, Universitätsklinikum Hamburg-Eppendorf
-
Alexander Schulte
- Klinik und Poliklinik für Neurochirurgie, Labor für Hirntumorbiologie, Universitätsklinikum Hamburg-Eppendorf
| Published: | May 13, 2014 |
|---|
Outline
Text
Objective: The contribution of intratumorally heterogeneous egfr amplification to glioblastoma biology is unclear. We present a unique EGFR-amplified, EGFRvIII-positive model system which recapitulates the clonal heterogeneity of primary glioblastoma tissue and allows analyzing the impact of EGFR amplification on tumor initiation, progression and response to EGFR-targeted therapy in a natively amplified background in vitro and in vivo.
Method: EGFR-amplified (BS153) or non-amplified (U87) glioblastoma cells were transduced with equivalent titers of three lentiviral vectors encoding red, green and blue fluorescent protein (RGB-marking) resulting in hundreds of individually colored cell clones in vitro based on their respective transduction ratio, allowing for tracking of individual cell clones in vitro and in vivo.
Results: When analyzed for their tumor-initiating capacity in a xenograft mouse model using NMRI/Foxn1nu-mice in vivo, BS153-derived tumors led to symptom development within eight weeks after implantation. The experimental tumors displayed an unusual, invasive morphology, probably attributable to EGFR overexpression, as opposed to the nodular in vivo growth pattern of non-amplified GBM cell lines such as U87MG. Strikingly, when RGB-marked BS153 cells were implanted into nude mice, tumors with only a limited number of colors developed, indicating that only a restricted number of cells have the ability to initiate tumors de novo. In contrast, RGB-marked U87MG glioblastoma cells gave rise to mixed multicolored tumors in vivo, indicating that a high proportion of the originally labeled cells engrafted in vivo and contributes to tumor formation. By contrast, selecting for properties that refer resistance to EGFR-targeted therapy, no selection for resistant clones occurred in vitro. Resistance was rather mediated by a change in gene expression pattern, highlighted by increased expression of the oncogenic EGFRvIII deletion variant.
Conclusions: Our results suggest an EGFR-associated cellular hierarchy for tumor initiation and progression in vivo. However, molecular determinants for in vivo-tumorigenicity appear to differ from those for acquired TKI resistance in vitro.
