The method of choice to knock-in large inserts via CRISPR

CRISPR/Cas9 gene editing is a revolutionary method used to study gene function by transcript silencing, knock-out, or activation. The knock-in of DNA fragments to endogenous genes of interest is another promising approach to study molecular pathways but is technically challenging. Many approaches have been suggested, but the proof of correct integration has often been relied on less convenient validation experiments. Within this work, we investigated homology-directed repair (HDR), non-homologous end joining (NHEJ), and PCRextension (PCRext) based approaches as three different methods to knock-in large DNA fragments (>1000 bp), and compared feasibility, cost effectiveness, and reliability. As a knock-in fragment, we used a fluorescent reporter sequence in order to directly assess successful integration by microscopy, subsequently proven by sequencing. For NHEJ and PCRext, we demonstrate that it is insufficient to rely on the fluorescent reporter due to false positive results. Both NHEJ and PCRext failed to reliably knock-in large DNA sequences, they were accompanied by massive generation of InDels driving the methodology cost-intensive and non-reliable. In contrast, combination of CRISPR/Cas9 and HDR revealed correct integration, proven by correct fluorescence of the subcellular localization and sequencing, and thus, corresponds to the method of choice for large fragment integration. Next to HEK293T, we demonstrate successful HDR based knock-in in human induced pluripotent stem cells (hiPSCs). Subsequent differentiation of gene-edited hiPSCs into cerebral organoids showed relevance of the approach to study subcellular protein localization and abundance in 3D tissue.