BIX-01294

Whilst there are many published references using this molecule in various cell-based studies, there is limited selectivity data disclosed for it beyond the closely associated family of proteins listed above.  In contrast, a closely structurally related molecule, UNC0638, has been published and has the above biochemical selectivity data plus data for wider selectivity. BIX-01294 should be used with caution because there is no published evidence to prove that it doesn't bind to other proteins, which could potentially confound cellular data.

BIX-01294 should be used with caution. I have several specific concerns about this molecule:

Selectivity over GLP is reportedly ~20x and >>20x for five related family members. Subsequent reports suggest it is only 6x more selective for GLP (doi: 10.1038/nchembio.599). Regardless, the suppression of H3K9me2 in G9a-/- cells suggests there is off-target activity.

Cellular activity is limited. At 4 uM (2x IC50), H3K9me2 is decreased only 20% along with very modest reduction in target gene expression (5%). At higher concentrations, the compound is toxic. More potent molecules for G9A have been reported that have a larger margin between cellular activity and cellular toxicity (DOI: 10.1021/jm200903z).

BIX-01294 is apparently tolerated when administered to mice intraperitonally or orally at 40-100 mg/kg. In one report, a more profound effect (9x) was seen with the analog TM2-115 at the same dose, suggesting that BIX-01294 has been superceded by a better molecule; however, a second report argues that TM2-115 is, in fact, less efficacious. (doi: 10.1073/pnas.1205414109, doi: 10.1128/AAC.04419-14). Several reports indicate that the cellular activity of BIX-01294 agrees with siRNA knockdown or CRISPR/Cas9 knockout of EHMT2, suggesting that the molecule can have value as a tool compound with appropriate biological confirmation.

Users should note selectivity experiments were performed against a limited panel of targets in the initial report. The lack of specificity data of BIX-01294 versus other targets (besides GLP, PRMT1, ESET, SE7/9, SUV39H1) should be considered when using this compound. Users of this compound should also note its relatively low potency (low micromolar). At the low micromolar compound concentrations needed to achieve a reduction in H3K9me2 levels, it is possible that other targets are being engaged in addition to G9a. Furthermore, this compound is toxic at higher concentrations, and there appears to be a tight window between achieving the desired target effects while avoiding toxicity.

Users should consider the development history of this compound. The compound was identified in a high-throughput screen (HTS) performed without detergent (according to Supplemental Information in the original manuscript), which is a common practice to mitigate the effect of chemical aggregates often encountered in these assays. This compound may not be a chemical aggregator, but -to my knowledge- this possibility has not been investigated. The original manuscript also did not report any compound optimization by medicinal chemistry or SAR. While not an absolute requirement, such efforts are usually required to generate high quality chemical probes. Additional evidence of meaningful target engagement such as X-ray crystallography, SPR, and/or cell-based engagement assays such as thermal shift or labeled compound would enhance confidence with respect to its use in chemical biology. Given its potency, cell-based efficacy (only 20% reduction in H3K9me2 levels at 4 uM), as well as unanswered questions about off-target effects, without additional data I would not recommend its use in model organisms at this time. The limited selectivity assessement of BIX-0194 within the HMT family and also outside the family restricts its usefulness. The compound is not very potent in vitro (low micromolar) and the cellular activity on H3K9me2 at 2x the reported IC50 is modest. Better characterized and more potent G9a inhibitors like A-366 are available and should be used instead.