CRISPR targeting the 5’ regulatory region of FOXM1 will also disrupt RHNO1

There is growing interest in targeting the transcription factor FOXM1 as a therapeutic strategy for cancer, including in epithelial ovarian cancer (EOC) (1). However, small molecule FOXM1 inhibitors have mostly been limited by pleiotropic effects or incomplete target engagement (2). Thus, the approach taken in a recent paper in Drug Development Research, to target FOXM1 using CRISPR deletion of a regulatory region, is a novel approach to disrupt this oncogene (3).

The authors showed that treatment of EOC cells with BET inhibitors downregulated FOXM1 expression, which implied that an enhancer region marked by H3K27Ac drives FOXM1 expression in EOC cells. In agreement, BET inhibitors were previously shown to repress FOXM1 expression (4). The authors mined ChIP-seq data to identify enhancers suitable for CRISPR deletion mediated FOXM1 repression. In Fig. 3A the authors presented ENCODE H3K27Ac data, showing a peak flanking the 5’ end of the FOXM1 gene and extending into the promoter. Visible in the figure is the fact that this peak also overlaps the 5’ end of a gene encoded on the opposite DNA strand, RHNO1. Critically, we have previously shown that FOXM1 and RHNO1 are bidirectional genes (BDGs) regulated by a bidirectional promoter (BDP) (5). In Fig 3B, the authors show their CRISPR deletion strategy while in Fig 3C they show PCR verification of genetic knockout. Curiously, the genomic coordinates provided in Fig 3B map to a region >100kB upstream of FOXM1, while the CRISPR guides presented in Fig 3C and Table 1 map to a distinct region, the sequences flanking exon 1 of FOXM1 and exon 1 of RHNO1. Therefore, the authors’ detection of the recombined allele in OVCAR3 cells is consistent with disruption of the first exons of both FOXM1 and RHNO1, along with the entire BDP, although the authors only analyzed the expression of FOXM1 in their study.

Using this new cell model, the authors demonstrate that the recombined OVCAR3 cells have attenuated oncogenic characteristics, including reduced clonogenic growth, cell proliferation, and wound healing (migratory) capacity in vitro, and reduced tumor growth in subcutaneous xenografts (3). While these effects are consistent with known functions of FOXM1 in EOC, they are also consistent with accumulating knowledge of RHNO1 (1, 6). We have shown that RHNO1 functions in EOC cell proliferation and clonogenic growth, while others showed that RHNO1 supports xenograft growth (7). Based on the known functions of FOXM1 and RHNO1, the newly generated OVCAR3 cell model should be evaluated to determine the extent to which the anti-cancer phenotypes observed are due to loss of FOXM1, RHNO1, or both. Importantly, we previously observed that in EOC cells the effects of FOXM1 and RHNO1 depletion together are greater than seen for either alone (5). Approaches such as genetic rescues may help to resolve the underlying mechanism of anti-cancer phenotypes in this model. More generally, the strategy employed by Chen et al. may be a novel means to simultaneously disrupt the oncogenic activity of BDGs, including FOXM1/RHNO1 but also other BDG pairs. Future research in this area, potentially taking advantage of nanoformulations suitable for CRISPR delivery, is worthwhile.