Genome editing of oncogenes with ZFNs and TALENs: caveats in nuclease design

Abstract

Background: Gene knockout technologies involving programmable nucleases have been used to create knockouts in several applications. Gene editing using Zinc-finger nucleases (ZFNs), Transcription activator like effectors (TALEs) and CRISPR/Cas systems has been used to create changes in the genome in order to make it non-functional. In the present study, we have looked into the possibility of using six fingered CompoZr ZFN pair to target the E6 gene of HPV 16 genome.

Methods: HPV 16^+ve cell lines; SiHa and CaSki were used for experiments. CompoZr ZFNs targeting E6 gene were designed and constructed by Sigma-Aldrich. TALENs targeting E6 and E7 genes were made using TALEN assembly kit. Gene editing was monitored by T7E1 mismatch nuclease and Nuclease resistance assays. Levels of E6 and E7 were further analyzed by RT-PCR, western blot as well as immunoflourescence analyses. To check if there is any interference due to methylation, cell lines were treated with sodium butyrate, and Nocodazole.

Results: Although ZFN editing activity in yeast based MEL-I assay was high, it yielded very low activity in tumor cell lines; only 6% editing in CaSki and negligible activity in SiHa cell lines. Though editing efficiency was better in CaSki, no significant reduction in E6 protein levels was observed in immunocytochemical analysis. Further, in silico analysis of DNA binding prediction revealed that some of the ZFN modules bound to sequence that did not match the target sequence. Hence, alternate ZFN pairs for E6 and E7 were not synthesized since no further active sites could be identified by in silico analyses. Then we designed TALENs to target E6 and E7 gene. TALENs designed to target E7 gene led to reduction of E7 levels in CaSki and SiHa cervical cancer cell lines. However, TALEN designed to target E6 gene did not yield any editing activity.

Conclusions: Our study highlights that designed nucleases intended to obtain bulk effect should have a reasonable editing activity which reflects phenotypically as well. Nucleases with low editing efficiency, intended for generation of knockout cell lines nucleases could be obtained by single cell cloning. This could serve as a criterion for designing ZFNs and TALENs.

Keywords: Cervical cancer; Gene editing; HPV; TALEN; Zinc-finger nucleases.

Fig. 1

CompoZr ZFNs designed to target E6 gene of HPV 16. a Schematic representation of ZFN targeting a region in exon 1 region of HPV 16-E6. b Represents the nucleotide sequence of the target region of E6. c An illustration of the designed ZFNs targeting E6 region containing six ZFs on either side with a spacer region of 5 nucleotides. d Graph depicts yeast MEL-I assay showing activity of CompoZr ZFNs (red bar). T7E1 editing showing ~ 6% editing in CaSki cell line (e) and no editing in SiHa cell line (f) upon ZFN treatment. Further, complementary nuclease resistance (NR) assay also yielded ~ 6% editing in CaSki cell line (g), while NR assay did not yield any editing for SiHa cell line (h)

Fig. 2

ZFN mediated editing indicated by 53bp1 (red spots) in CaSki cells. A–D Control cells with vector alone (green) showed no red spots (53bp1). E–H Treated cells with CompoZr ZFNs flag tagged (FITC-green) showed red spots indicating double strand breaks. Magnification ×600

Fig. 3

Expression analysis of E6 in CaSki cells showed no significant difference between control and treated groups. a RT-PCR data indicated that although full length E6 levels were decreased in treated, E6* spliced variant (300 bp) levels remained the same in both control and treated. This was further validated by immunocytochemical analysis for E6 proteins (b–g). ZFN treated CaSki cells showed no difference in E6 (red) levels after editing (e–g) when compared to control cells (b–d). Western blot analysis corroborated with the immunocytochemical analysis indicating no significant difference in E6 levels before and after treatment (h). Magnification ×600

Fig. 4

B1H screens of C2H2-ZF domains. B1H screens indicated that there was difference in the predicted binding site and the target sequence of CompoZr ZFN 1 and 2

Fig. 5

In silico ZFN target site prediction for E6 gene using ZiFiT tools did not yield suitable target sites. E6 gene was screened for possible ZFN binding regions using a CoDa specific zinc fingers, b OPEN validated zinc fingers, c Sangamo validated modules, d TOOLGEN validated naturally occurring ZFN modules, e Barbas group of validated ZFN modules and f ZIF Predict

Fig. 6

TALEN targeting E7 in SiHa and Caski cells showed reduction in E7 levels. a Schematic representation of TALEN targeting E7. b T7E1 assay showing cut products in both SiHa and CaSki cell lines with TALEN treatment. Further, RT-PCR revealed a reduction in E7 levels in SiHa (c) and CaSki (d) cell lines

Fig. 7

TALEN mediated editing in SiHa and CaSki cell lines. a, b SiHa control, c, d TALEN treated SiHa cells showed single spots (green; arrows) indicating the presence of 53bp1, e, f CaSki control, g, h TALEN treated CaSki cell showed multiple spots (green; arrows) indicating the presence of DNA double strand breaks. Magnification ×600

Fig. 8

TALEN-mediated editing resulted in the reduction of E7 protein in both SiHa and CaSki cells. Immunocytochemical analysis revealed that levels of E7 protein was reduced in TALEN-treated (C, D) when compared to control (A, B) SiHa cells. Similarly, a reduction in E7 levels was observed in CaSki cells also after TALEN treatment (G, H) than in its respective control cells (E, F). This was further confirmed by Western blot analysis which showed a significant reduction in the levels of E7 after TALEN treatment in both the cell lines (I). Magnification-×600