Anti-VEGFR2 driven nuclear translocation of VEGFR2 and acquired malignant hallmarks are mutation dependent in glioblastoma

Abstract

Objective: Anti-angiogenic therapies (AATs), targeting VEGF-VEGFR pathways, are being used as an adjuvant to normalize glioblastoma (GBM) vasculature. Unexpectedly, clinical trials have witnessed transient therapeutic effect followed by aggressive tumor recurrence. In pre-clinical studies, targeting VEGFR2 with vatalanib, increased GBM growth under hypoxic microenvironment. There is limited understanding of these unanticipated results. Here, we investigated tumor cell associated phenotypes in response to VEGFR2 blockade.

Methods: Human U251 cells were orthotopically implanted in mice (day 0) and were treated with vehicle or vatalanib on day 8. Tumor specimens were collected for immunohistochemistry and protein array. Nuclear translocation of VEGFR2 was analyzed through IHC and western blot. In vitro studies were performed in U251 (p53 and EGFR mutated) and U87 (p53 and EGFR wildtype) cells following vehicle or vatalanib treatments under normoxia (21% O₂) and hypoxia (1% O₂). Proliferation, cell cycle and apoptosis assays were done to analyze tumor cell phenotypes after treatments.

Results: Vatalanib treated animals displayed distinct patterns of VEGFR2 translocation into nuclear compartment of U251 tumor cells. In vitro studies suggest that vatalanib significantly induced nuclear translocation of VEGFR2, characterized in chromatin bound fraction, especially in U251 tumor cells grown under normoxia and hypoxia. Anti-VEGFR2 driven nuclear translocation of VEGFR2 was associated with increased cell cycle and proliferation, decreased apoptosis, and displayed increased invasiveness in U251 compared to U87 cells.

Conclusions: Study suggests that AAT- induced molecular and phenotypic alterations in tumor cells are associated with mutation status and are responsible for aggressive tumor growth. Therefore, mutation status of the tumor in GBM patients should be taken in to consideration before applying targeted therapy to overcome unwanted effects.

Keywords: Glioblastoma; VEGFR2; anti-angiogenic therapy; hypoxia; mutation; nuclear translocation; resistance; tumor cells.

Figure 1. Anti-VEGFR2 induced nuclear translocation of VEGFR2 in GBM

(A) Immunohistochemical analysis on vehicle and vatalanib treated tumor tissues (n=3). Left panels showing tumor areas in whole brain (2.5 X). Middle panel showing VEGFR2 expression in enlarged (40 X) tumor sections. Vehicle treated tumors displayed mostly membrane bound VEGFR2 (yellow arrows). Vatalanib treated tumors displayed increased nuclear VEGFR2 (red arrows). (B) Quantitative data showing significant increased nuclear VEGFR2 in vatalanib treated tumors compared to vehicle. (C) Western blot data showing protein expression of HIF1α, total VEGFR2 and VEGFR2 expression in cellular compartments of U251 and U87 in response to vehicle and vatalanib (10μM) treatment under normoxic and hypoxic conditions. (D) Immunohistochemical analysis on cultured U251 and U87 tumor-cell smears showing nuclear expression of VEGFR2 along with vehicle vs. vatalanib treatment as well as normoxia vs. hypoxia. Shown is one of the two experiments performed. Quantitative data is expressed in mean ±SD and ***P<0.001.

Figure 2. Vatalanib treatment induced cell cycle and proliferations of GBM cells

(A) Cell cycle analysis of U251 and U87 cells showing effect of vatalanib treatment on G0/G1, S and G2/M phases under normoxia and hypoxia. In U251 cells, vatalanib treatment showed modest effect on G0/G1, S and G2/M phases under normoxia compared to vehicle group. Vatalanib significantly increased G2/M phase under hypoxia condition compared to vehicle (5.1% vs 14.1%). In U87 cells, vatalanib treatment resulted into G0/G1 arrest under normoxia compared to vehicle (50.6% vs 61.2%). Vatalanib treatment showed modest effect on G0/G1, S and G2/M phases under hypoxia compared to vehicle. (B) Western blot showing vatalanib induced ERK expression in U251 and U87 cells under normoxia and hypoxia. In U251 cells, vatalanib treatment increased phospho-ERK expression both in normoxia and hypoxia. No change was seen in phospho-ERK expression in U87 cells. Shown is one of the two experiments performed.

Figure 3. Vatalanib treatment decreased apoptosis of GBM cells

(A) Flow cytometry data showing annexin V staining on U251 and U87 cells. In U251 cells, normoxia group displayed increased viability (21.8% vs 18.2%), and decreased dead cells (8.2% vs 12.2%) with no change in apoptosis (69.8% vs 69.5%) with vatalanib compared to vehicle, respectively. Hypoxia resulted in increased viable cells (51.3% vs 37.9%), decreased dead cells (2.4% vs 3.3%) and decreased apoptotic cells (46% vs 58.6%) with vatalanib compared to vehicle, respectively (left and right panels). In U87 cells, vatalanib treatment under normoxia increased live cells (19.8% vs 15.4%), increased dead cells (13.3% vs 6.6%) and decreased apoptotic cells (66.6% vs 77.8%) compared to vehicle, respectively. Similarly, vatalanib under hypoxia increased U87 cell viability (49.1% vs 37.5%), decreased apoptotic cells (43.4% vs 56.7%) and increased dead cells (7% vs 5.7%), respectively (left and right panels). (B) Western blot data showing decreased cleaved PAPR expression in both U251 and U87 cells under hypoxia compared to normoxia. Shown is one of the two experiments performed.

Figure 4. Vatalanib treatment induced invasion and migration of U251 cells

(A) Membrane protein array data showing significant overexpression of chemokines and their receptors (EGF-EGFR, TGFB1-TGFBR2, PDGFAA-PDGFRA, MMP2 and MMP9) in tumor conditioned medium following vatalanib treatment compared to vehicle, especially under hypoxia. (B) Western blot data showing increased phospho-P38 and ID1 protein expression in U251 cells both under normoxia and hypoxia, following vatalanib treatment compared to vehicle. In U87 cells, vatalanib decreased phospho-P38 expression under normoxia compared to vehicle group. 1D1 expression was increased after vatalanib treatment under normoxia compared to vehicle group. However, hypoxia significantly decreased ID1 expression and did not change with vatalanib. Overall, p-AKT was upregulated in hypoxia compared to normoxia in both U251 and U87. Interestingly, vatalanib decreased p-AKT compared to vehicle under normoxia in U251 cells. Vatalanib increased p-AKT compared to vehicle under hypoxia in U87 cells. (C) Vatalanib treated tumor showing increased CD44+ invasive tumor cells by immunofluorescence staining. Invasive front in vatalanib group was characterized by the CD44+ tumor cells, which were migrating away from the tumor periphery (yellow arrows). Shown is one of the two experiments performed. Quantitative data is expressed in mean ±SD. * P<0.05 and **P<0.01.