Proceedings of The Physiological Society

Europhysiology 2018 (London, UK) (2018) Proc Physiol Soc 41, PCA021

Poster Communications

HIF2a mediates hypoxia-induced expression of the Wilms tumor protein WT1 in Kelly neuroblastoma cells

L. Catanese1, L. K. Sciesielski2, K. M. Kirschner1, H. Scholz1, K. Krueger1

1. Inst. of Vegetative Physiology, Charité - Universitätsmedizin Berlin, Berlin, Germany. 2. Department of Neonatology, Charité - Universitätsmedizin Berlin, Berlin, Germany.


Neuroblastoma is a malignant childhood tumor arising from the ganglia of the peripheral sympathetic nervous system. A striking feature of this cancer is its biological and clinical heterogeneity. Tissue hypoxia favors an undifferentiated tumor phenotype with poor prognosis. Recent findings indicate that the clinical outcome of a subgroup of neuroblastomas, which lack amplification of the MYCN oncogene negatively correlates with the expression level of the Wilms tumor gene WT1. To obtain novel insights into the molecular mechanisms underlying neuroblastoma growth, we analyzed the expression of WT1 in human neuroblastoma cell lines with the MYCN amplification (IMR-32, SK-N-BE(2), Kelly) and in non-MYCN amplified (SK-N-AS, SH-EP1, SH-SY-5Y) neuroblastoma lines. In addition to normoxic conditions (21% O2), experiments were also performed at 1% O2 to mimic tumor hypoxia. In normoxia (21% O2), robust WT1 protein levels were detected in 4 out of the 6 investigated neuroblastoma cell lines. WT1 expression did not correlate with the status of MYCN-amplification. Exposure to 1% O2 (24 h) increased WT1 mRNA and protein more than 10-fold in Kelly cells (p<0.01, n=6). Hypoxia stimulated WT1 expression also in SK-N-BE(2) cells but not in the other cell lines tested. WT1 mRNA and protein levels in Kelly cells at 21% O2 were enhanced approximately 9-fold by stabilization of hypoxia-inducible factors (HIFs) with dimethyloxalylglycine (DMOG) and 2,2'-Dipyridyl (2,2'-DP, 100 µM each). Silencing of HIF2a, but not HIF1a, by RNA interference significantly reduced WT1 protein and mRNA levels at 1% O2. Since transfection with siRNAs did not completely abolish HIF expression in hypoxia, we generated Kelly cell lines with inactivation of either HIF1a or HIF2a using CRISPR/Cas9. Mutations in the coding regions of both genes were identified by sequencing of PCR-amplified genomic DNA, and gene inactivation was confirmed by the lack of HIF1a and HIF2a proteins in the hypoxic knockout cell lines. HIF2a deficiency caused a loss of WT1 expression in Kelly cells exposed to either 1% O2 or 2,2'-DP (21% O2). By contrast, hypoxia-induced expression of WT1 was preserved in Kelly cells with HIF1a knockout. Finally, silencing of WT1 by RNA interference significantly reduced the migration of SKNAS neuroblastoma cells in transwell filter assays. These findings demonstrate that hypoxia stimulates WT1 expression in Kelly cells via HIF2a. In conclusion, up-regulation of WT1 expression in hypoxia may contribute to the malignant growth phenotype by stimulating tumor cell migration in a subgroup of neuroblastomas.

Where applicable, experiments conform with Society ethical requirements