Biophysical characterisation of the Bcl-x pre-mRNA and binding specificity of the ellipticine derivative GQC-05: Implication for alternative splicing regulation

Abstract

The BCL2L1 gene expresses two isoforms of Bcl-x protein via the use of either of two alternative 5′ splice sites (5′ss) in exon 2. These proteins have antagonistic actions, Bcl-XL being anti-apoptotic and Bcl-XS pro-apoptotic. In a number of cancers the Bcl-XL isoform is over-expressed, resulting in cancer cell survival and growth, so switching splicing to the Xs isoform could have therapeutic benefits. We have previously proposed that a putative G-quadruplex (G4) exists downstream of the XS 5′ss and shown that the ellipticine derivative GQC-05, a previously identified DNA G4-specific ligand, induces an increase in the XS/XL ratio both in vitro and in cells. Here, we demonstrate that this G4 forms in vitro and that the structure is stabilised in the presence of GQC-05. We also show that GQC-05 binds RNA non-specifically in buffer conditions, but selectively to the Bcl-x G4 in the presence of nuclear extract, highlighting the limitations of biophysical measurements taken outside of a functional environment. We also demonstrate that GQC-05 is able to shift the equilibrium between competing G4 and duplex structures towards the G4 conformation, leading to an increase in accessibility of the XS 5′ss, supporting our previous model on the mechanism of action of GQC-05.

Alternative splicing of BCL-x is controlled by RBM25 binding to a G-quadruplex in BCL-x pre-mRNA

“…The capacity of GQ-2 fragment to form rG4 structure in vitro has recently been demonstrated by Bhogadia et al. through CD, 1 H nuclear magnetic resonance and gel electrophoresis experiments ( 35 ). Here, we additionally assessed the capacity of GQ-2 and its variants to form G4 structures in vitro using TDS and UV melting experiments.…”

“…In addition, thanks to FOLDeR, a new strategy based on footprinting of long 7-deazaguanine-substituted RNA, the formation of two of these rG4 was confirmed in vitro under functional conditions: one (GQ-2) lying in close proximity with the 5′ss leading to the short (alternative) pro-apoptotic Bcl-xS isoform (xS 5′ss) while the other (GQ-5) is overlapping the 5′ss leading to the long (canonical) anti-apoptotic Bcl-xL isoform (xL 5′ss) ( 33 ). In addition, the ability of GQ-2 to adopt a G4 conformation has been recently confirmed by various biophysical experiments ( 34 , 35 ).…”

“…The first hypothesis is supported by the effects of G4 ligands such as PhenDC3, which are likely to increase the lifetime and/or the fraction of GQ-2 rG4 at the expense of alternative secondary RNA structures, thereby enhancing the binding of RBM25. At the same time, in favour of the second hypothesis, recent in vitro data ( 35 ) showed that the stabilizing effect of a G4 ligand (GQC-05, an ellipticin derivative) on GQ-2 rG4 could be observed only in the presence of nuclear extract. This suggests that the existence of this rG4 in cellulo requires the presence of a nuclear factor, which may be an RNA binding protein such as RBM25, that would promote its formation and thereby its recognition and further stabilization by a G4 ligand.…”