Improved cell-penetrating zinc-finger nuclease proteins for precision genome engineering

Abstract

Safe, efficient, and broadly applicable methods for delivering site-specific nucleases into cells are needed in order for targeted genome editing to reach its full potential for basic research and medicine. We previously reported that zinc-finger nuclease (ZFN) proteins have the innate capacity to cross cell membranes and induce genome modification via their direct application to human cells. Here, we show that incorporation of tandem nuclear localization signal (NLS) repeats into the ZFN protein backbone enhances cell permeability nearly 13-fold and that single administration of multi-NLS ZFN proteins leads to genome modification rates of up to 26% in CD4(+) T cells and 17% in CD34(+) hematopoietic stem/progenitor cells. In addition, we show that multi-NLS ZFN proteins attenuate off-target effects and that codelivery of ZFN protein pairs facilitates dual gene modification frequencies of 20-30% in CD4(+) T cells. These results illustrate the applicability of ZFN protein delivery for precision genome engineering.

Figure 1

Tandem NLS repeats enhance ZFN protein activity. (a) Diagrams of one- to five-NLS ZFN proteins. Green and white boxes indicate NLS and poly-His domains, respectively. (b) Schematic representation of the HEK293 EGFP reporter system used to evaluate multi-NLS ZFN protein activity. “CCR5-R” indicates the “right” CCR5 ZFN protein binding sites. (c) Percentage of EGFP-positive reporter cells measured by flow cytometry following treatment with increasing concentrations of one- to five-NLS ZFN protein. (d) Percentage of EGFP-positive reporter cells measured by flow cytometry following one to three consecutive treatments with 0.5 μmol/l one- to five-NLS ZFN protein. (e) Percentage of FITC-positive HEK293 cells measured by flow cytometry following treatment with 1 μmol/l fluorescein-conjugated one- to five-NLS ZFN proteins for 1 hour. “Mock” indicates cells treated with serum-free medium. Bars represent ± SD (n = 3). *P < 0.05; **P < 0.01; ***P < 0.001 by t-test.

Figure 2

Modification of endogenous genes by direct delivery of multi-NLS ZFN proteins. (a,b) Frequency of endogenous CCR5 gene modification in K562 and Jurkat cells treated with (a) 0.5 or 4 μmol/l one- to five-NLS ZFN proteins targeting the CCR5 gene or (b) increasing concentrations of four-NLS ZFN proteins targeting the CCR5 gene. Modification determined by the Surveyor nuclease assay. “M” indicates cells treated with serum-free medium.

Figure 3

Modification of endogenous genes in primary cells and stem cells by direct delivery of multi-NLS ZFN proteins. (a,b) Frequency of endogenous CCR5 gene modification in (a) stimulated human CD4⁺ T cell treated with increasing amounts of one- to five-NLS ZFN proteins targeting the CCR5 gene or (b, left) hematopoietic stem/progenitor cells (HSPCs) and (b, right) induced pluripotent stem cells (iPSCs) treated with 4.0 μmol/l native (one-NLS) or multi-NLS ZFN proteins targeting the CCR5 gene. “M” indicates cells treated with serum-free medium. Gene modification determined by the Surveyor nuclease assay. Bars represent ± SD (n = 3). *P < 0.05, **P < 0.01, ***P < 0.001 by t-test.

Figure 4

Cleavage specificity of multi-NLS ZFN proteins. Surveyor nuclease analysis of the CCR5, PGC, FBXL11, ZCCHC14, and C3orf59 loci in K562 cells treated with 4 μmol/l one- to five-NLS ZFN proteins targeting the CCR5 gene. Gene modification and the ratio between off-target to on-target cleavage are denoted. “DNA” indicates cells nucleofected with 1 μg each of the “left” and “right” CCR5 one-NLS ZFN expression vectors. “M” indicates cells treated with serum-free medium. Arrows indicate anticipated cleavage products.

Figure 5

Toxicity induced by multi-NLS ZFN proteins. Viability of stimulated CD4⁺ T cells treated once with increasing concentrations of one- to five-NLS ZFN proteins targeting the CCR5 gene. Proliferation was measured 2 days after protein treatments. Data normalized to cells treated with serum-free medium. Bars represent ± SEM (n = 3). *P < 0.05; **P < 0.01; ***P < 0.001 by t-test.

Figure 6

Modification of the CCR5 and CXCR4 genes in CD4⁺ T cells by direct delivery of multi-NLS ZFN protein pairs. (a) Frequency of endogenous CXCR4 gene modification in HeLa, K562, and Jurkat cells treated with 4 μmol/l one-, three-, and four-NLS ZFN proteins targeting the CXCR4 gene. (b) Frequency of endogenous CCR5 and CXCR4 gene modification in stimulated CD4⁺ T cells treated simultaneously with 2 μmol/l each three-NLS ZFN proteins targeting the CCR5 and CXCR4 genes. Gene modification was determined by Surveyor nuclease assay. Arrows indicate expected cleavage products. “M” indicates cells treated with serum-free medium and “D2” and “D5” indicate genomic DNA isolated from CD4⁺ T cells at 2 and 5 days after ZFN protein treatments, respectively.