AROs knock down KRT14 expression in an hnRNPA1 specific manner.
To test whether AROs can knock down gene expression in a targeted manner, we created a 23 base-pair antisense RNA molecule with homology to exon 7 of the KRT14 gene and fused an hnRNPA1 recruiting loop to either the 5’ end, the 3’ end or both sides of the targeting oligonucleotide. KRT14 was chosen as a knockdown target because it has been implicated as a driver of the epithelial-mesenchymal transition in numerous cancers, (Bilandzic et al., 2019, p. 14; Papafotiou et al., 2016, p. 14) and Exon 7 was chosen due to its small size (47 base-pairs) and because it contains several motifs for ASF/SF2 exonic splice enhancer binding, a regulatory element that hnRNPA1 has been shown to repress (Cartegni et al., 2003; Kashima et al., 2007). As controls, we also tested an oligonucleotide containing only the 23 bp targeting domain (i.e. an “antisense only” control with no A1 recruiting loop) and an ARO with a scrambled targeting domain (Fig 2A), which respectively served to determine whether the targeting oligonucleotide or hnRNPA1-recruiting loop alone influenced KRT14 expression. Human keratinocytes, HaCat cells, were chosen as a model cell line for initial experiments due to their robust expression of KRT14.
Because hnRNPA1 binds cooperatively (Jean-Philippe et al., 2013), we hypothesized that we would see the greatest knockdown of gene expression with the ARO containing two flanking hnRNPA1 recruiting loops (ARL-Oligo-ARL). Consistent with our expectations, we observed a 56% knockdown of KRT14 gene expression in ARL-Oligo-ARL compared to the carrier control (P <= 0.001), and a 53% knockdown of expression compared with the targeting oligo only (Oligo) control (P <= 0.01) indicating that the hnRNPA1 recruiting loop was required for knockdown to occur (Figure 2B). We also saw significant knockdown of the ARL-Oligo (P <= 0.05) and Oligo-ARL (P <= 0.05) experimental conditions (18% and 29% respectively), when compared to the targeting oligo only (Oligo) control. Interestingly, we observed a small (15%) but statistically significant difference between the targeting oligo only control (Oligo) and the scrambled targeting oligo control attached to an hnRNPA1 recruiting loop (scrOligo-ARL) indicating that the presence of the ARL alone may have a minor knockdown effect on KRT14. Alternatively, it is possible that the scrambled targeting oligo has off-target effects influencing the expression of other transcripts which affect the expression of KRT14.
To test whether the effects of the AROs were hnRNPA1 dependent, each condition was repeated in the presence of two validated siRNAs designed to knockdown hnRNPA1 (Figure 2B). Concurrent loss of hnRNPA1 with expression of ARL-Oligo, Oligo-ARL, and ARL-Oligo-ARL caused significant and complete loss of previously seen knockdown effects on KRT14 transcript (P <= 0.05, P <= 0.01 and P <= 0.01 respectively). In contrast, differences between the control constructs, scrOligo-ARL and Oligo, and their hnRNPA1 siRNA knocked down counterparts were not significant.
Taken together, our results demonstrate that AROs targeting KRT14 knock down mRNA expression in a targeted and hnRNPA1 dependent manner.