GoldyTALEN Vectors with Improved Efficiency for Golden Gate TALEN Assembly

TALEN-mediated genome editing has widespread utility in many systems for reverse genetic approaches. GoldyTALEN, ¹ a Golden Gate–compatible platform for TALEN assembly, ² was previously optimized by Bedell et al. ³ We found that the pT3Ts-goldyTALEN (pT3TsgT) Golden Gate vector is prone to recombination, leading to an increased number of false-positive colonies during the final TALEN assembly step. Here we report an updated vector, called pKT3Ts-goldyTALEN (pKT3TsgT), which was modified to prevent recombination and allow for selection with the more stable antibiotic kanamycin. The pKT3TsgT vector is compatible with the Golden Gate TALEN assembly protocol and produces TALENs with the same assembly efficiency and in vivo activity as the previous pT3TsgT vector. These modifications in pKT3TS-gT allowing for efficient assembly during automated high throughput TALEN production.

The GoldyTALEN scaffold was previously cloned into the mRNA expression vector pT3TS ⁴ to create pT3TS-gT. ³ We found that initial TALEN assemblies with pT3TS-gT were nearly 100% efficient, with 10/10 white colonies representing complete TALEN clones. However, with newer preparations of pT3TS-gT propagated in bacteria, the number of false-positive clones (white colonies) lacking TAL repeats routinely reached 90–95%, lowering the efficiency to 5–10%. Examination of the pT3TS-gT vector sequence revealed two lacZ sequences on opposing strands flanking the TAL RVD cloning site that could recombine and remove the RVD restriction sites and the FokI sequence necessary for TALEN assembly (Fig. 1A). We observed both blue and white colonies after pT3TS-gT transformation, indicating that the plasmid stock contained a population of vectors lacking the lacZ selection gene (Fig. 1C). We re-engineered the vector to remove the secondary lacZ sequence (Fig. 1B and Supplementary Materials and Methods; Supplementary Data are available online at www.liebertpub.com/hum) and prevent recombination (Fig. 1D).

OPEN IN VIEWER

Figure 1.

Comparison of recombination between the original pT3Ts-goldyTALEN (pT3TsgT) and the modified pKT3Ts-goldyTALEN (pKT3TsgT) vectors. (A) The original pT3TsgT vector has two lacZ sequences on opposite strands, allowing for recombination and excision of the RVD cut sites and the FokI gene. (B) The modified pKT3TsgT vector lacks the second lacZ site and has a backbone that contains the kanamycin resistance gene. (C) Transformation of pT3TsgT and plating on LB-Xgal plates with the appropriate antibiotic produces 5 times as many white colonies as blue. (D) Plating pkT3TsgT on LB-Xgal plates results in only blue colonies. (E) Comparison of targeting efficiency between TALENs assembled in the pT3TsgT and pKT3TsgT vectors. Each TALEN pair targeting either rb1 or cdh5 resulted in 100% biallelic inactivation after injection into individual embryos. +, digested; −, undigested PCR amplicons.

To verify that the new backbone did not alter TALEN mutagenesis activity, we compared TALEN mRNAs targeting the zebrafish rb1 and cdh5 genes produced with the pT3TS-gT and pKT3TS-gT vectors. For both rb1 and cdh5 genes, TALEN injection resulted in PCR amplicons that contained a restriction site in the targeted region and were almost completely resistant to restriction digestion (Fig. 1E), suggesting that insertion or deletion alleles were induced at frequencies that approach 100%. In contrast, control PCR amplification products from uninjected embryos were fully digested. These data indicate that TALENs generated from either vector will target mutations with equal efficacy in vivo. In summary, the modified pKT3TS-gT vector avoids recombination of the parental plasmid in bacteria, allowing for more efficient and reliable generation of TALENs for genetic studies in any TALEN-compatible system and aiding small laboratories in adopting this technology. We anticipate that the increased vector stability of pKT3TS-gT will also facilitate efficient one-step TALEN assembly. ⁵