Insertion of YFP at the Arabidopsis AFL1 and P5CS1 loci shows the potential, and potential limitations, of gene targeting for functional analysis of stress-related genes

Transgenic expression of a tagged protein under control of its native promoter is the typical strategy to investigate tissue-specific accumulation and subcellular localization of a protein of interest. In some cases, difficulties in promoter identification, altered promoter function when placed in a new genomic context, and variability between transgenic lines can complicate such experiments. Crispr/CAS9-enabled homologous recombination to insert a tag in frame with an endogenous gene could circumvent these problems. Apart from work at a few loci used to test gene targeting (GT) protocols, there have been relatively few reports of GT to knock-in protein tags in Arabidopsis thaliana , and little data of whether expression of the knock-in allele accurately reflects that of the wild-type gene. We investigated these questions using a two-stage GT procedure where a starter line expressing hyper-accurate CAS9 ( hypaCAS9 ) under control of the Egg Cell1.1 ( EC1.1 ) promoter was transformed with an mCitrine (YFP) donor sequence flanked by homology arms for homologous recombination at the N- or C-terminal ends of the stress-responsive genes At14a-Like 1 ( AFL1 ) and Δ1 -pyrroline-5-carboxylate synthetase 1 ( P5CS1 ). We observed high frequency GT to generate in frame fusions of YFP at either the N- or C-terminus of AFL1 and at the N-terminus of P5CS1 . AFL1-YFP was produced from the knock-in locus; however, the N-terminal knock-ins disrupted AFL1 and P5CS1 expression. These results show that the combination of EC1.1 promoter and hypaCAS9 along with selection of a starter line with strong hypaCAS9 expression produced good GT frequencies, indicating that this approach can be feasible to knock-in or replace sequence at many loci. However, we also offer a cautionary tale that insertions generated by GT can interfere with gene expression for reasons which are yet unclear. Further refinement of this approach is a potentially useful tool to produce tagged proteins from their endogenous genomic loci. ### Competing Interest Statement The authors have declared no competing interest.