FLASH assembly of TALENs for high-throughput genome editing

Abstract

Engineered transcription activator–like effector nucleases (TALENs) have shown promise as facile and broadly applicable genome editing tools. However, no publicly available high-throughput method for constructing TALENs has been published, and large-scale assessments of the success rate and targeting range of the technology remain lacking. Here we describe the fast ligation-based automatable solid-phase high-throughput (FLASH) system, a rapid and cost-effective method for large-scale assembly of TALENs. We tested 48 FLASH-assembled TALEN pairs in a human cell–based EGFP reporter system and found that all 48 possessed efficient gene-modification activities. We also used FLASH to assemble TALENs for 96 endogenous human genes implicated in cancer and/or epigenetic regulation and found that 84 pairs were able to efficiently introduce targeted alterations. Our results establish the robustness of TALEN technology and demonstrate that FLASH facilitates high-throughput genome editing at a scale not currently possible with other genome modification technologies.

Figure 1. Graphical overview of the FLASH assembly method

(a) Archive of 376 TALE repeat encoding plasmids required to practice FLASH. Plasmids encoding one, two, three, and four TALE repeats (colored rectangles) harboring various RVDs (represented by two upper-case letters within the rectangles) were constructed as described in Methods. (b) Schematic overview of the FLASH assembly process. A DNA fragment encoding a single TALE repeat and labeled on its 5′ end with biotin (blue oval) is initially ligated to a second DNA fragment encoding four specific TALE repeats and then attached to a streptavidin-coated magnetic bead (orange sphere). Additional DNA fragments encoding pre-assembled TALE repeats are ligated until an array of the desired length is assembled. The DNA fragment encoding the full-length TALE repeat array is then cleaved from the bead by restriction digestion.

Figure 2. Gene-disruption activities of 48 FLASH-assembled TALENs targeted to EGFP

(a) Activities of 48 TALEN pairs and four ZFN pairs in the EGFP gene-disruption assay. Percentages of EGFP-disrupted cells as measured 2 and 5 days following transfection of U2OS cells bearing a chromosomally integrated EGFP reporter gene with nuclease-encoding plasmids are shown. Target sites for the 48 TALEN pairs are shown in Supplementary Table 3. The four ZFN pairs were made by OPEN and have been previously described. Mean percent disruption of EGFP and standard error of the mean from three independent transfections are shown. All 48 TALEN pairs and all four ZFN pairs tested induced statistically significant (p<0.05) increases in EGFP-disrupted cells relative to control transfected cells on both day 2 and day 5. Asterisks indicate TALEN and ZFN pairs for which statistically significant differences exist between values of EGFP-disrupted cells on day 2 and day 5 (p<0.05). (b) Mean EGFP-disruption activities from (a) grouped by length of the TALENs. (c) Ratio of mean percent EGFP disruption values from day 2 to day 5. Ratios were calculated for groups of each length TALEN using the data from (b). Values greater than 1 indicate a decrease in the average percentage of EGFP-disrupted cells at day 5 relative to day 2. (d) Ratio of mean tdTomato-positive cells on day 2 and day 5 grouped by various lengths of TALENs. tdTomato-encoding control plasmids were transfected together with nuclease-encoding plasmids on day 0.

Figure 3. Computationally-derived design guidelines do not show statistically significant correlation with TALEN activity levels

Each graph plots the mean EGFP-disruption activity of each of our 48 EGFP-targeted TALEN pairs against the number of half-sites within each target site that fail to meet one of four previously described design guidelines (Guidelines 2, 3, 4, or 5; see Supplementary Discussion) or against the total number of guideline violations in each target site (Total Guideline Violations). Plots are shown for EGFP-disruption activities measured on day 2 and day 5 post-transfection. Guideline violation data are from Supplementary Table 3. Correlation p-values are shown for each plot.

Figure 4. DNA sequences and frequencies of FLASH TALEN-induced mutations at endogenous human genes

For each endogenous gene target, the wild-type (WT) sequence is shown at the top with the TALEN target half-sites highlighted in yellow and the translation start codon of the gene (ATG) underlined and highlighted in bold type. Deletions are indicated by red dashes against a grey background and insertions by lowercase letters against a light blue background. The sizes of the insertions (+) or deletions (▵) are indicated to the right of each mutated site. The number of times that each mutant was isolated is shown in parentheses. Mutation frequencies are calculated as the number of mutants identified divided by the total number of sequences analyzed. Note that for several of the genes, we also identified larger deletions that extend beyond the sequences of the TALEN target sites. For each gene, sequencing was performed with the same genomic DNA used to perform the corresponding T7EI assays shown in Supplementary Figure 3.

Figure 4. DNA sequences and frequencies of FLASH TALEN-induced mutations at endogenous human genes

For each endogenous gene target, the wild-type (WT) sequence is shown at the top with the TALEN target half-sites highlighted in yellow and the translation start codon of the gene (ATG) underlined and highlighted in bold type. Deletions are indicated by red dashes against a grey background and insertions by lowercase letters against a light blue background. The sizes of the insertions (+) or deletions (▵) are indicated to the right of each mutated site. The number of times that each mutant was isolated is shown in parentheses. Mutation frequencies are calculated as the number of mutants identified divided by the total number of sequences analyzed. Note that for several of the genes, we also identified larger deletions that extend beyond the sequences of the TALEN target sites. For each gene, sequencing was performed with the same genomic DNA used to perform the corresponding T7EI assays shown in Supplementary Figure 3.

Figure 4. DNA sequences and frequencies of FLASH TALEN-induced mutations at endogenous human genes

For each endogenous gene target, the wild-type (WT) sequence is shown at the top with the TALEN target half-sites highlighted in yellow and the translation start codon of the gene (ATG) underlined and highlighted in bold type. Deletions are indicated by red dashes against a grey background and insertions by lowercase letters against a light blue background. The sizes of the insertions (+) or deletions (▵) are indicated to the right of each mutated site. The number of times that each mutant was isolated is shown in parentheses. Mutation frequencies are calculated as the number of mutants identified divided by the total number of sequences analyzed. Note that for several of the genes, we also identified larger deletions that extend beyond the sequences of the TALEN target sites. For each gene, sequencing was performed with the same genomic DNA used to perform the corresponding T7EI assays shown in Supplementary Figure 3.

Figure 4. DNA sequences and frequencies of FLASH TALEN-induced mutations at endogenous human genes

For each endogenous gene target, the wild-type (WT) sequence is shown at the top with the TALEN target half-sites highlighted in yellow and the translation start codon of the gene (ATG) underlined and highlighted in bold type. Deletions are indicated by red dashes against a grey background and insertions by lowercase letters against a light blue background. The sizes of the insertions (+) or deletions (▵) are indicated to the right of each mutated site. The number of times that each mutant was isolated is shown in parentheses. Mutation frequencies are calculated as the number of mutants identified divided by the total number of sequences analyzed. Note that for several of the genes, we also identified larger deletions that extend beyond the sequences of the TALEN target sites. For each gene, sequencing was performed with the same genomic DNA used to perform the corresponding T7EI assays shown in Supplementary Figure 3.