It sounds like you really just need to make a conditional knock-in mouse. Maybe with Crispr technology, this can be done fairly efficiently (see this paper for example: http://www.cell.com/retrieve/pii/S0092867413010167). Overall, the project seems a bit risky though, since full Elk1 knock-outs might only exhibit mild phenotypes (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC303347/). It is an interesting question, however, to try to block the phosphorylation of one specific kinase substrate and not just block all of the kinase activity. I wonder how you would translate any findings into a therapy, though.

Thanks for the response, Mark. Crispr may provide the means to address such a question. I feel that the behavioral assays I have in mind for these mutant mice might be sensitive enough to elucidate an Elk-1 dependent effect --- re: Cesari et al. 2004, there are likely developmental compensatory mechanisms, perhaps exhibiting functional redundancy as the authors suggest, for Elk-1 by other ternary complex factors to continue transcription during development and throughout experience to lead to a seemingly normal adult brain (although the gamut of experience-dependent gene expression is long and this study is certainly not exhaustive, but a good initial characterization nonetheless). The goal of the experiment I have in mind is to elucidate the contribution of certain "players" in the signaling cascade that leads to a particular behavioral effect our lab has previously observed, so I feel selective inhibition will be necessary to achieve this. Again, thank you for the suggestion and the hasty response!