Targeting the Oncogenic Long Non-coding RNA SLNCR1 by Blocking Its Sequence-Specific Binding to the Androgen Receptor

Schmidt et al., Cell Reports, 2020 asked how the long non-coding RNA SLNCR1 holds on to the androgen receptor, and whether that grip can be prised apart with an oligonucleotide. Using chemical probing of RNA structure alongside binding assays, the group showed that the receptor's N-terminal domain reads a pyrimidine-rich motif that sits in an unstructured stretch of the transcript. The motif turned out to be predictive rather than incidental, because the same rule found a receptor binding motif in a second long non-coding RNA, HOXA11-AS-203.

Having found the contact point, the authors went after it with oligonucleotides supplied by AUM BioTech, between 21 and 28 nucleotides long. Two were AUMblock sdASOs antisense to the receptor binding sequence in SLNCR1, which pair with it and hide it. The other two were sense-strand copies of the same sequence, which work the other way round by binding the receptor protein itself. The chemistry was chosen so that the oligonucleotides do not support cleavage of what they bind, so the transcript survives and only the interaction is blocked. In a cell-free binding assay all four blocked the receptor from binding the transcript.

In A375 melanoma cells the oligonucleotides were delivered gymnotically, added to the medium with no transfection reagent and no cell-transducing agent. Overexpressing SLNCR1 raised invasion through Matrigel, and every one of the four oligonucleotides brought it back to the level of the vector and scramble control. The amount of SLNCR1 in the cells did not change, which is what a steric block is supposed to do, and the rise in MMP9 that the transcript causes was attenuated. The paper's own framing is that blocking a protein and long non-coding RNA contact, rather than destroying the transcript, is a route worth pursuing where the transcript itself is not the problem.