Evaluating the Effect of Chromosomal Context on Zinc Finger Nuclease Efficiency

By: Scott Bahr, Laura Cortner, Sara Ladley, Trissa Borgschulte, CHOZN® Platform Development Team, SAFC/Sigma-Aldrich, St Louis, MO 63103 USA, scott.bahr@sial.com


Zinc Finger Nuclease (ZFN) technology has provided researchers with a tool for integrating exogenous sequences into most cell lines or genomes in a precise manner. Using current methods, the efficiency of targeted integration (TI) into the host genome is generally low and is highly dependent on the ZFN activity at the genomic locus of interest. It is unknown if the ZFN binding and cutting efficiency is more dependent on the nucleotide recognition sequence or the chromosomal context in which the sequence is located.

We have taken a highly efficient ZFN pair (hAAVS1) from human studies and introduced the exogenous ZFN DNA recognition sequence into the Chinese Hamster Ovary (CHO) genome in an attempt to improve the efficiency of targeted integration. A “Landing Pad” comprised of human AAVS1 genomic sequence encompassing the ZFN recognition sequence has been integrated into the CHO genome at 3 separate loci to determine if the ZFN’s will work across species and if the cutting efficiency is affected by chromosomal context. The results of this study will help us to improve the overall efficiency of TI by using Landing Pads, particularly for genomic targets in which suitable ZFN’s may not be available.

3 CHO Loci were chosen for this study based on previous gene expression studies. Rosa26 and Neu3 show consistent but low levels of expression while Site #1 appears to have no known coding sequence. Additionally, Rosa26 and Site#1 were chosen as potential safe harbor sites in CHO. The ZFN cutting efficiency at the endogenous CHO loci Rosa26, Site #1 and Neu3 are approximately 15%, 30% and 40% respectively. Based on other studies the cutting efficiency of human AAVS1 ZFN’s was as high as 50% depending on the human cell line used. The hAAVS1 Landing Pad was integrated at CHO Rosa26, Site #1, and Neu3 via homologous recombination.

Clones carrying the exogenous hAAVS1 Landing Pads at Rosa26, Site #1 and Neu3 were transfected with hAAVS1 ZFN’s and the cutting efficiency was measured. We found that the human AAVS1 ZFN’s were able to successfully cut at their recognition sequence in the Landing Pad at all 3 CHO loci to varying degrees. These results indicate that the chromosomal context of the ZFN recognition sequence has an effect on cutting efficiency. This study shows that TI can be performed with Landing Pads across species with high efficiency and provide researchers with additional tools for cell line engineering.



Figure 1. Workflow of Generating Clones with Landing Pads

Plasmid donors were designed for each of the 3 CHO Loci to facilitate TI of the hAAVS1 Landing Pad. Single cell clones were isolated and screened by jPCR. Positive clones were banked and sequence confirmed. Once we had Landing Pads at all 3 loci the cell lines were transfected with the same hAAVS1 ZFN’s (mRNA) and relative cutting efficiency was determined by Cel1.


Figure 2. Schematic hAAVS1 Landing Pad Integration

The Landing Pad includes 100-200 bp of the hAAVS1 genomic sequence and encompasses the ZFN recognition Sequence. The Landing Pad is flanked by homology arms specific to the 3 endogenous CHO loci to facilitate TI.



Figure 3. Confirmation of Landing Pad Integration at 3 CHO Loci

Clones were screened by PCR spanning the integration sites using primers specific to Rosa26, Site#1 and Neu3. Positive clones were further sequence confirmed for Landing Pad integration. The gel above shows PCR at Site #1 of a WT clone, a heterozygous clone and a clone with integration of the Landing Pad in both alleles.


Figure 4. Determination of ZFN activity at 3 endogenous CHO Loci

A. 3 CHO genomic loci were transfected with the best available ZFN designs and cutting efficiency was measured by Cel1 Assay or direct sequencing of Indels. ZFN induced double stranded breaks in the genome are often repaired by Non-Homologous End Joining (NHEJ) which creates indels in the ZFN target sequence. PCR amplified Indels are annealed and create mismatches in the PCR product at the ZFN target site. Cel1 Nuclease cleaves these mismatched products and creates secondary bands in addition to the WT PCR product. The percentage of cleaved amplicons correlates with ZFN activity. B. The level of ZFN activity at CHO Rosa26 was calculated based Cel1 Assay while ZFN activity at Site#1 and Neu3 was determined by direct sequencing of Indels which is more accurate.  


Table 2. hAAVS1 ZFN cutting efficiency at a Landing Pad in 3 CHO loci

3 cell lines containing the hAAVS1 Landing Pad at the 3 separate loci were transfected with the hAAVS1 ZFN to compare activity. ZFN efficiency at each loci was measured by Cel1 Assay or direct sequencing of Indels in PCR amplicons. We see successful ZFN activity at all 3 loci but with varying efficiency. **The Landing Pad integration at Neu3 locus caused phenotypic changes in the cell growth and viability following transfection which may explain low ZFN activity. The cutting efficiency at the Landing Pad at Neu3 was calculated by sequencing Indels and not Cel1 Assay  

Conclusions and Discussion

  • This work describes the first successful use of a ZFN across species and at multiple genomic locations within the same species.
  • We describe a novel approach to improving TI efficiency in CHO by first introducing an exogenous landing pad for future targeting.
  • Chromosomal context and chromatin structure likely has a major effect on ZFN binding and cutting efficiency.
  • The hAAVS1 ZFN cutting efficiency varies between 3 genomic loci even though the recognition sequence is identical.
  • This study shows that ZFN encoding Landing Pads could be beneficial to improving TI efficiencies. Additionally, Landing Pads with multiple ZFN target sequences can be used to facilitate sequential integrations at a loci.


Carolyn Weigand - sequencing support
Greg Davis - support with the hAAVS1 ZFN’s


Related Links