|Home | About | Journals | Submit | Contact Us | Français|
Transcription activator-like effectors (TALEs) are a class of naturally occurring DNA binding proteins found in the plant pathogen Xanthomonas sp. The DNA binding domain of each TALE consists of tandem 34-amino acid repeat modules that can be rearranged according to a simple cipher to target new DNA sequences. Customized TALEs can be used for a wide variety of genome engineering applications, including transcriptional modulation and genome editing. Here we describe a toolbox for rapid construction of custom TALE transcription factors (TALE-TFs) and nucleases (TALENs) using a hierarchical ligation procedure. This toolbox facilitates affordable and rapid construction of custom TALE-TFs and TALENs within one week and can be easily scaled up to construct TALEs for multiple targets in parallel. We also provide details for testing the activity in mammalian cells of custom TALE-TFs and TALENs using, respectively, qRT-PCR and Surveyor nuclease. The TALE toolbox described here will enable a broad range of biological applications.
Systematic reverse engineering of the functional architecture of the mammalian genome requires the ability to perform precise perturbations on gene sequences and transcription levels. Tools capable of facilitating targeted genome editing and transcription modulation are essential for elucidating the genetic and epigenetic basis of diverse biological functions and diseases. Recent discovery of the transcription activator-like effector (TALE) code1, 2 has enabled the generation of custom TALE DNA binding domains with programmable specificity3–12. When coupled to effector domains, customized TALEs provide a promising platform for achieving a wide variety of targeted genome manipulations3–5, 8, 11, 13, 14. Previously, we reported efficient construction of TALEs with customized DNA binding domains for activating endogenous genes in the mammalian genome3. Here we describe an improved protocol for rapid construction of customized TALEs and methods to apply these TALEs to achieve endogenous transcriptional activation3–5, 8 and site-specific genome editing4, 7, 9, 11–15. Investigators should be able to use this protocol to construct TALEs for targets of their choice in less than one week.
TALEs are natural bacterial effector proteins used by Xanthomonas sp. to modulate gene transcription in host plants to facilitate bacterial colonization16, 17. The central region of the protein contains tandem repeats of 34 amino acids sequences (termed monomers) that are required for DNA recognition and binding18–21 (Fig. 1a). Naturally occurring TALEs have been found to have a variable number of monomers, ranging from 1.5 to 33.5 (ref. 16). Although the sequence of each monomer is highly conserved, they differ primarily in two positions termed the repeat variable diresidues (RVDs, 12th and 13th positions). Recent reports have found that the identity of these two residues determines the nucleotide binding specificity of each TALE repeat and a simple cipher specifies the target base of each RVD (NI = A, HD = C, NG = T, NN = G or A)1, 2. Thus, each monomer targets one nucleotide and the linear sequence of monomers in a TALE specifies the target DNA sequence in the 5′ to 3′ orientation. The natural TALE binding sites within plant genomes always begin with a thymine1, 2, which is presumably specified by a cryptic signal within the non-repetitive N-terminus of TALEs. The tandem repeat DNA binding domain always ends with a half length repeat (0.5 repeat, Fig. 1a). Therefore, the length of DNA sequence being targeted is equal to the number of full repeat monomers plus two.
For targeted gene insertion and knockout, there are several techniques that have been used widely in the past, such as homologous gene targeting22–24, transposases25, 26, site-specific recombinases27, meganucleases28, and integrating viral vectors29, 30. However most of these tools target a preferred DNA sequence and cannot be easily engineered to function at non-canonical DNA target sites. The most promising, programmable DNA-binding domain has been the artificial zinc finger (ZF) technology, which enables arrays of ZF modules to be assembled into a tandem array and target novel DNA binding sites in the genome. Each finger module in a ZF array targets three DNA bases31, 32. In comparison, TALE DNA binding monomers target single nucleotides and are much more modular than ZF modules. For instance, when two independent ZF modules are assembled into a new array, the resulting target site cannot be easily predicted based on the known binding sites for the individual finger modules. Perhaps the biggest caveat of ZFs is that most of the intellectual property surrounding the ZF technology platform is proprietary and expensive (>$10k per target site). A public effort for ZF technology development also exists through the Zinc Finger Consortium but the publicly available ZF modules can only target a subset of the 64 possible trinucleotide combinations33–35. TALEs theoretically can target any sequence and have already been deployed in many organisms with impressive success (see Table 1). Although TALEs seem superior in many ways, zinc fingers have a much longer track record in DNA-targeting applications32, including their use in human clinical trials36. Despite their relatively recent development, early results with TALEs have been promising and it seems that they can be applied in the same way as zinc fingers for many DNA-targeting applications (e.g. transcriptional modulator3–5, 8, nuclease4, 7, 9, 11–15, recombinase37–39, transposase40, 41).
Due to the repetitive nature of TALEs, construction of the DNA-binding monomers can be difficult. Previously, we and other groups have used a hierarchical ligation strategy to overcome the difficulty of assembling the monomers into ordered multimer arrays, taking advantage of degeneracy in codons surrounding the monomer junction and Type IIs restriction enzymes3, 6–10. In this protocol, we employ the same basic strategy that we previously used3 to construct TALE-TFs to modulate transcription of endogenous human genes. We have further improved the TALE assembly system with a few optimizations, including maximizing the dissimilarity of ligation adaptors to minimize misligations and combining separate digest and ligation steps into single Golden Gate42–44 reactions. Briefly, we first amplify each nucleotide-specific monomer sequence with ligation adaptors that uniquely specify the monomer position within the TALE tandem repeats. Once this monomer library is produced, it can conveniently be re-used for the assembly of many TALEs. For each TALE desired, the appropriate monomers are first ligated into hexamers, which are then amplified via PCR. Then, a second Golden Gate digestion-ligation with the appropriate TALE cloning backbone (Fig. 1b) yields a fully-assembled, sequence-specific TALE. The backbone contains a ccdB negative selection cassette flanked by the TALE N- and C-termini, which is replaced by the tandem repeat DNA-binding domain when the TALE has been successfully constructed. ccdB selects against cells transformed with an empty backbone, therefore yielding clones with tandem repeats inserted7.
Assemblies of monomeric DNA binding domains can be inserted into the appropriate TALE transcription factor (TALE-TF) or TALE nuclease (TALEN) cloning backbones to construct customized TALE-TFs and TALENs. TALE-TFs are constructed by replacing the natural activation domain within the TALE C-term with the synthetic transcription activation domain VP64 (ref. 3) (Fig. 1c). By targeting a binding site upstream of the transcription start site, TALE-TFs recruit the transcription complex in a site-specific manner and initiate gene transcription. TALENs are constructed by fusing a C-term truncation (+63aa) of the TALE DNA binding domain4 with the non-specific FokI endonuclease catalytic domain (Fig. 1d). The +63aa C-term truncation has also been shown to function as the minimal C-term sufficient for transcriptional modulation3. TALENs form dimers through binding to two target sequences separated by ~17 bases. Between the pair of binding sites, the FokI catalytic domains dimerize and function as molecular scissors by introducing double-strand breaks (DSBs) (Fig 1d). Normally, DSBs are repaired by the non-homologous end-joining45 (NHEJ) pathway, resulting in small deletions and functional gene knock-out. Alternatively, TALEN-mediated DSBs can stimulate homologous recombination, enabling site-specific insertion of an exogenous donor DNA template4, 13.
We also present a short procedure for verifying correct TALE assembly: using colony PCR to verify the correct insert length followed by DNA sequencing. With our cloning procedure, we routinely achieve high efficiency (correct length) and high accuracy (correct sequence). The cloning procedure is modular in several ways: We can construct TALEs to target DNA sequences of different lengths and the protocol is the same for producing either TALE-TFs or TALENs. The backbone vectors can be modified with different promoters to achieve cell-type specific expression.
Our protocol includes functional assays for evaluating TALE-TF and TALEN activity in human cells. This step is important because we have observed some variability in TALE activity on the endogenous genome, possibly due to epigenetic repression and/or inaccessible chromatin at certain loci. For TALE-TFs, we perform quantitative reverse-transcription polymerase chain reaction (qRT-PCR) to quantify changes in gene expression. For TALENs, we use the Surveyor mutation detection assay (i.e. the base-mismatch cleaving endonuclease Cel2) to quantify NHEJ. Although these assays are standard and have already been described elsewhere46, 47, we feel that the functional characterization is integral to TALE production and therefore have presented it here with the assembly procedure. Other functional assays such as plasmid-based reporter constructs3, 7, restriction sites destroyed by NHEJ48, or other enzymes that detect DNA mismatch49 may also be used to validate TALE activity.
Our protocol (Fig 2) begins with the generation of a monomer library, which takes one day and can be re-used for building many TALEs. Using the monomer library, several TALEs can be constructed in a single day with an additional two days for transformation and sequence verification. To assess TALE function on the endogenous genome, we take ~3 days to go from mammalian cell transfection to qRT-PCR or Surveyor results.
A number of TALE assembly procedures have described the use of Golden-Gate cloning to construct customized TALE DNA binding domains3, 6–10. These methods rely on the use of a large collection of plasmids (typically over 50 plasmids) encoding repeat monomers and intermediate cloning vectors. Our PCR-based approach requires significantly less initial plasmid preparation, as our monomer library can be amplified on one 96-well PCR plate, and facilitates more rapid construction of custom TALEs. Plasmid-based amplification has a much lower mutation/error rate but, in our experience, the combination of a high-fidelity polymerase and the short length of the monomer template (~100 nt) results in accurate assembly. For building similar length TALEs to those presented in this protocol, the plasmid-based approaches also require an additional transformation and colony selection that extends the time needed to build TALEs. Thus, these alternative assembly protocols require a greater time investment both upfront (for monomer library preparation) and on a recurring basis (for each new TALE). For laboratories seeking to produce TALEs quickly, our protocol requires only a few hours to prepare a complete monomer library and less than a day to proceed from monomers to the final transformation into bacteria.
There are a few important limitations with the TALE technology. Although the RVD cipher is known, it is still not well understood why different TALEs designed according to the same cipher act on their target sites in the native genome with different levels of activity. It is possible that there are yet unknown sequence dependencies for efficient binding or site-specific constraints (e.g. chromatin state) that are responsible for differences in functional activity. Therefore we suggest constructing at least 2 or 3 TALE-TFs or TALEN pairs for each target locus. Also, it is possible that engineered TALEs can have off-target effects – binding unintended genomic loci – which can be difficult to detect without additional functional assays at these loci. Given the relatively early state of TALE technology development, these issues remain to be addressed in a conclusive manner.
The programmable nature of TALEs allows for a virtually arbitrary selection of target DNA binding sites. As previously reported, the N-terminus of the TALE requires that the target site begin with a thymine nucleotide. For TALE-TFs, we have been successful targeting 14 to 20 bp sequences within 200 bp of the transcription start site (Fig. 1c). It can be advantageous to select a longer sequence to reduce off-target activation, as it is known from reporter activation assays that TALEs interact less efficiently with targets contain more than one mismatching base. In our assembly protocol, we describe ligation of 18 monomers into a backbone containing a nucleotide-specific final 0.5 monomer; combined with the initial thymine requirement, this yields a total sequence specificity of 20 nucleotides. Specifically, the TALE-TF binding site takes the form 5′-TN19-3′. When selecting TALE-TF targeting sites for modulating endogenous gene transcription, we recommend selecting multiple target sites within the proximal promoter region (can target either the sense or antisense strand), as epigenetic and local chromatin dynamics might impede TALE binding. Larger TALEs might be beneficial for TALE-TFs targeting genes with less unique regions upstream of their transcription start site.
Since TALENs function as dimers, a pair of TALENs, referred to as the left and right TALENs, need to be designed to target a given site in the genome. The left and right TALENs target sequences on opposite strands of DNA (Fig. 1d). As with TALE-TF, we design each TALEN to target a 20 base pair sequence. TALENs are engineered as a fusion of the TALE DNA-binding domain and a monomeric FokI catalytic domain. To facilitate FokI dimerization, the left and right TALEN target sites are chosen with a spacing of ~14–20 bases. Therefore, for a pair of TALENs, each targeting 20 base pair sequences, the complete target site should have the form 5′-TN19N14–20N19A-3′, where the left TALEN targets 5′-TN19-3′ and the right TALEN targets the antisense strand of 5′-N19A-3′ (N = A, G, T, or C). TALENs should have fewer off-target effects due to the dimerization requirement for the FokI nuclease, although no significant off-target effects have been observed in limited sequencing verifications13. Because DSB formation only occurs if the spacer between the left and right TALEN binding sites (Fig. 1d) is ~14–20 bases, nuclease activity is restricted to genomic sites with both the specific sequences of the left TALEN and the right TALEN with this small range of spacing distances between those sites. These constraints should greatly reduce potential off-target effects.
To ensure that all synthesized TALEs are transcribed at a similar level, all of the monomers have been optimized to share identical DNA sequences except in the variable di-residues – and are codon-optimized for expression in human cells (see Supplementary Data 1). This should minimize any difference in translation due to codon availability.
Synthesis of monomeric TALE DNA binding domains in a precise order is challenging due to their highly repetitive nature. Previously3, we took advantage of codon redundancy at the junctions between neighboring monomers and devised a hierarchical ligation strategy to construct ordered assemblies of multiple monomers. In this protocol, we describe a similar strategy but with several important improvements that make the procedure easier, more flexible, and more reliable (Fig. 3).
In our initial protocol3, the digestion and ligation steps were carried out separately with an intervening DNA purification step. This improved protocol adopts the powerful Golden Gate cloning technique42–44, requiring less hands-on time and resulting in a more efficient reaction. The Golden Gate procedure involves combining the restriction enzyme and ligase together in a single reaction with a mutually compatible buffer. The reaction is cycled between optimal temperatures for digestion and ligation. Golden Gate digestion-ligation capitalizes on Type IIs restriction enzymes, for which the recognition sequence is spatially separated from where the cut is made. During a Golden Gate reaction, the correctly ligated products no longer contain restriction enzyme recognition sites and cannot be further digested. In this manner, Golden Gate drives the reaction toward the correct ligation product as the number of cycles of digestion and ligation increases.
For the hierarchical ligation steps, we have optimized our previous cloning strategy for faster TALE production. The improved design takes advantage of a circularization step that allows only properly assembled hexameric intermediates to be preserved (Fig. 3). Correctly ligated hexamers consist of six monomers ligated together in a closed circle, and incomplete ligation products are left as linear DNA. After this ligation step, an exonuclease degrades all non-circular DNA, leaving intact only the complete circular hexamers. Without circularization and exonuclease treatment, the correct ligation product would need to be gel purified before proceeding. The combination of Golden Gate digestion-ligation and circularization reduces the overall hands-on time required for TALE assembly.
Each monomer in the tandem repeat must have its position uniquely specified. The monomer primers are designed to add ligation adaptors that enforce this positioning. Our protocol uses a hierarchical ligation strategy: For the 18mer tandem repeat, we first ligate monomers into hexamers. Then, we ligate three hexamers together to form the 18mer. By breaking down the assembly into two steps, we do not need unique ligation junctions for each monomer in the 18mer. Instead the same set of ligation junctions internal to each hexamer are re-used in all three hexamers (first ligation step), whereas unique (external) ligation junctions are used to flank each hexamer (second ligation step). As shown in Figure 4, the internal primers used to amplify the monomers within each hexamer are the same, but the external primers differ between the hexamers. By re-using the same internal primers between different hexamers, our protocol minimizes the number of primers necessary for monomer amplification.
As a negative control for Golden Gate assembly, we recommend performing a separate reaction with only the TALE-TF or TALEN backbone. Transformation of this negative control should result in few or no colonies due to the omission of the tandem repeats and resulting re-ligation of the toxic ccdB insert. After completing the TALE cloning, we use colony PCR or restriction digests to screen for correct length clones. For the final verification of proper assembly, we sequence the entire length of the tandem repeats. Due to limits in Sanger sequencing read length, other TALE assembly protocols have difficulty sequencing the entire tandem repeat region7, 9, 10. The similarity of the monomers within the region makes primer annealing to specific monomers impossible. We have overcome this problem by slightly modifying the codon usage at the 5′ end of monomer 7 to create a unique annealing site so that a TALE with a 18mer DNA binding array can be verified through a combination of three staggered sequencing reads. Specifically, during the monomer amplification, the codons for the first 5 amino acids in monomer 7 are mutated via PCR to use different but synonymous codons, creating a unique priming site without changing the encoded TALE protein. This modification allows each hexamer in the 18mer to be sequenced with a separate sequencing read and requires only a standard read length of ~700 bp for complete sequence verification. For TALEs containing more than 18 full monomer repeats, we introduced a third unique priming site for sequencing at the 3′ end of the 18th monomer using a similar approach. For construction of TALEs containing up to 24 full monomers with the entire tandem repeat region easily sequenced, see Box 1.
In the main protocol, we present a hierarchical ligation strategy for the construction of TALEs that contain 18 full monomer repeats; however the general approach can be easily adapted to construct TALEs of any length. These TALEs containing 18 full repeat monomers bind to 20 bp DNA sequences, where the first and last bases are specified by the N-terminus and the 0.5 repeat, respectively (Fig 1a). We chose this length because, empirically, we have observed that 20 bp sequences tend to be unique within the human genome. Nevertheless, for different species (eg. with larger or more repetitive genomes) or for repetitive regions within the human genome, it can be advantageous to construct longer or shorter TALEs. For certain genomic loci, it might also be difficult to identify TALEN target sites that satisfy the spacing constraints when the binding sites for both left and right TALENs are restricted to 20 bp sequences.
Our protocol is easily modified for the construction of TALEs containing up to 24 full monomer repeats by changing the order in which particular primers are used during the preparation of the monomer library plate (as described in Procedure Steps 1 – 9). All other steps remain essentially the same. A plate of monomer amplification primers (similar to Figure 4) can be prepared for building TALEs with 24 full monomer repeats, which bind to 26 bp DNA sequences, as illustrated below. In this case, a fourth circular hexamer, corresponding to monomers 19 through 24, is also built and treated identically as the other three circular hexamers (1 – 6, 7 – 12, and 13 – 18).
For building shorter TALEs, only a single change to monomer amplification is needed: The final monomer should be amplified with the Ex-R4 reverse primer. For example, to build TALEs with 17 monomers instead of 18, the monomers templates (NI, NG, NN, HD) should be amplified with the forward/reverse primer combination In-F5/Ex-R4. Note that during gel purification (Procedure Step 20) the desired PCR amplicon is a pentamer containing monomers 13 – 17 and will run faster than the hexamers (1 – 6, 7 – 12). After purification, ensure that the pentameric and hexameric intermediates are used at an equimolar ratio in the final Golden Gate digestion-ligation.
For TALE-TFs, qRT-PCR quantitatively measures the increase in transcription driven by the TALE-TF. For TALENs, the Surveyor assay provides a functional validation of TALEN cutting and quantifies the cutting efficiency of a particular pair of TALENs. These assays should be performed in the same cell type as intended for the TALE application, as TALE efficacy can vary between cell types, presumably due to differences in chromatin state or epigenetic modifications.
For qRT-PCR, we use commercially-available probes to measure increased transcription of the TALE-TF-targeted gene. For most genes in the human or mouse genomes, specific probes can be purchased (e.g. TaqMan Gene Expression Probes from Applied Biosystems). There are a wide variety of qRT-PCR protocols and, although we describe one of them here, others can be substituted. For example, a more economical option is to design custom, transcript-specific primers (e.g. with NCBI Primer-BLAST) and use a standard fluorescent dye to detect amplified dsDNA (e.g. SYBR Green).
For Surveyor, we follow the recommendations given by the assay manufacturer when designing specific primers for genomic PCR. We typically design primers that are ~30 nucleotides long and with melting temperatures of ~65 °C. The primers should flank the TALEN target site and generate an amplicon of ~300–800 bp with the TALEN target site near the middle. During the design, we also check to make sure the primers are specific over the intended genome using NCBI Primer-BLAST (http://www.ncbi.nlm.nih.gov/tools/primer-blast/). Before using the primers for Surveyor, the primers and specific PCR cycling parameters should be tested to ensure that amplification results in a single clean band. In difficult cases where a single band product cannot be achieved, it is acceptable to gel extract the correct length band before proceeding with heteroduplex re-annealing and Surveyor nuclease digest.
CRITICAL Standard Taq polymerase, which lacks 3′-5′ exonuclease proofreading activity, has lower fidelity and can lead to errors in the final assembled TALE. Herculase II is a high-fidelity polymerase (equivalent fidelity to Pfu) that produces high yields of PCR product with minimal optimization. Other high-fidelity polymerases may be substituted.
CRITICAL MinElute columns should be stored at 4°C until use.
CRITICAL Do not substitute the more commonly-used T4 ligase. T7 ligase has 1000-fold higher activity on sticky ends than blunt ends and higher overall activity than commercially available concentrated T4 ligases.
CRITICAL Since Surveyor assay is sensitive to single-base mismatches, it is important to use only a high-fidelity polymerase. Other high-fidelity polymerases can be substituted; refer to the Surveyor manual for PCR buffer compatibility details.
CRITICAL The Surveyor assay includes the Cel2 base-mismatch nuclease. Alternatives include the Cel1, T7, mung bean, and S1 nucleases50, 51. Of these, Cel1 has been applied extensively for mutation detection52–54 and established protocols are available for its purification52, 54.
CRITICAL Programmable temperature stepping is needed for the TALEN (Surveyor) functional assay. Other steps only require a PCR-capable thermocycler.
Dilute in distilled water to 1x working solution for casting agarose gels and as a buffer for gel electrophoresis. Buffer can be stored at room temperature indefinitely.
Dilute 100x bovine serum albumin (BSA, supplied with BsaI-HF) to 10x concentration and store at −20 °C for at least 1 year in 20 ul aliquots.
Divide 10 mM ATP into 50 ul aliquots and store at −20 °C for up to 1 year; avoid repeated freeze-thaw cycles.
Prepare 10 mM DTT solution in distilled water and store 20 ul aliquots at −70 °C for up to 2 years; for each reaction, use a new aliquot since DTT is easily oxidized.
For culture of HEK293FT cells, prepare D10 culture medium by supplementing DMEM with 1X GlutaMAX and 10% fetal bovine serum. As indicated in the protocol, this medium can also be supplemented with 1X penicillin-streptomycin. D10 medium can be made in advance and stored at 4 °C for up to 1 month.
|Monomer template plasmid (5 ng/ul)||2 ul||50 pg/ul|
|100 mM dNTP (25 mM each)||2 ul||1 mM|
|5X Herculase II PCR buffer||40 ul||1x|
|20 uM primer mix (10 uM forward primer and 10 uM reverse primers from Step 1)||4 ul||200 nM|
|Herculase II Fusion polymerase||2 ul|
|Distilled water||150 ul|
|Total||200 ul (for 2 reactions)|
|1||95°C, 2 min|
|2–31||95°C, 20 s||60°C, 20 s||72°C, 10 s|
|32||72°C, 3 min|
CRITICAL STEP Before eluting the DNA, let the 96-well column plate air dry, preferably at 37°C, for 30 minutes on a clean Kimwipe so that all residual ethanol has enough time to evaporate.
CRITICAL STEP For subsequent digestion and ligation reactions, it is important that all monomers are at equimolar concentrations.
PAUSE POINT Amplified monomers can be stored at −20 °C for several months and can be reused for assembling additional TALEs.
CRITICAL STEP Pay close attention when pipetting the monomers; it is very easy to accidentally pipette from the wrong well during this step.
|Esp3I (BsmBI) 5 U/ul||0.75 ul||0.375 U/ul|
|Tango Buffer 10X||1 ul||1x|
|Dithiothreitol (DTT) 10 mM||1 ul||1 mM|
|T7 Ligase 3000 U/ul||0.25 ul||75 U/ul|
|ATP 10 mM||1 ul
|6 monomers||6 × 1 ul|
CRITICAL STEP Dithiothreitol (DTT) is easily oxidized in air. It should be freshly made or thawed from aliquots stored at −70 °C and used immediately.
|1–15||37°C, 5||20°C, 5|
|Hold at 4°C.||min||min|
PAUSE POINT This reaction can be left to run overnight.
|PlasmidSafe DNAse 10U/ul||1 ul||0.66 U/ul|
|Plasmid-Safe 10X Reaction Buffer||1 ul||1x|
|ATP 10 mM||1 ul
|Golden gate reaction from Step 14||7 ul|
PAUSE POINT After completion, the reaction can be frozen and continued later. The circular DNA should be stable for at least a week.
|100 mM dNTP (25 mM each)||0.5 ul||1 mM|
|5X Herculase II reaction buffer||10 ul||1x|
|10 uM each Hex-F and Hex-R primers||1 ul||200 nM|
|Herculase II Fusion DNA polymerase||0.5 ul||1x|
|Distilled water||37 ul
|PlasmidSafe-treated hexamer from Step 17||1 ul|
|1||95°C, 2 min|
|2–36||95°C, 20 s||60°C, 20 s||72°C, 30 s|
|37||72°C, 3 min|
CRITICAL STEP Avoid any cross-contamination by ethanol sterilization of work surfaces, razor blades, etc. during the gel extraction and between each individual band excision.
! CAUTION Wear appropriate personal protective equipment, including a facemask, when performing gel stabs to minimize risks associated with prolonged light or mutagenic DNA dye exposure.
|Component||TALE||Negative control||Final concentration|
|TALE backbone vector (100 ng/ul)||1 ul||1 ul||10 ng/ul|
|BsaI-HF (20 U/ul)||0.75 ul||0.75 ul||1.5 U/ul|
|10x NEBuffer 4||1 ul||1 ul||1x|
|10x Bovine serum albumin||1 ul||1 ul||1x|
|ATP 10 mM||1 ul||1 ul||1 mM|
|T7 Ligase (3000 U/ul)||0.25 ul||0.25 ul||75 U/ul|
|5 ul||5 ul|
|3 purified hexamers (20 ng/ul)||3 ul (1 ul each)||2 ng/ul each|
|Distilled water||2 ul||5 ul|
|Total||10 ul||10 ul|
CRITICAL STEP As a negative control, set up a separate reaction omitting the purified hexamers (i.e. including only the TALEN or TALE-TF backbone).
|1–20||37°C, 5 min||20°C, 5 min|
|21||80°C, 20 min|
PAUSE POINT Ligation products can be frozen at −20 °C and stored at least one month for transformation into bacteria at a later time.
|Colony suspension from Step 31||1 ul|
|100 mM dNTP (25 mM each)||0.25 ul||1 mM|
|10x Taq-B polymerase buffer||2.5 ul||1x|
|10 uM each TALE-Seq-F1 and TALE-Seq-R1 primers||0.25 ul||100 nM|
|Taq-B polymerase (5 U/ul)||0.1 ul||0.02 U/ul|
|Distilled water||20.9 ul|
|1||94°C, 3 min|
|2–31||94°C, 30 s||60°C, 30 s||68°C, 2 min|
|32||68°C, 5 min|
CRITICAL STEP: The DNA concentration of the TALE plasmids should be quantified to guarantee that an accurate amount of TALE DNA will be used during the transfection.
i) Mix 4 μg of TALE-TF plasmid DNA with 250 μl of Opti-MEM medium. Include controls (e.g. RFP plasmid or mock transfection) to monitor transfection efficiency and cell health respectively.
i) Mix 2 μg of the Left and 2 μg of the Right TALEN (Figure 1d) plasmid DNA with 250 μl of Opti-MEM medium. Control transfections should be done by omitting one or both of the TALENs. Also include controls (e.g. an RFP plasmid or mock transfection) to monitor transfection efficiency and cell health respectively. For all transfections, make sure the total amount of DNA transfected is the same across conditions – when omitting one or both TALENs, supplement with empty vector DNA to maintain the same total DNA amount.
CRITICAL STEP Make sure the complex is thoroughly mixed. Insufficient mixing results in lower transfection efficiency.
PAUSE POINT The transfection complex will remain stable for 6 hours at room temperature.
CRITICAL STEP If incubation beyond 48 hours is needed, change the culture medium with fresh D10 supplemented with antibiotics on a daily basis. This will not affect the transfection efficiency.
|1||68°C, 15 min|
|2||95°C, 8 min|
|gDNA from Step 46Aii||0.5 ul|
|100 mM dNTP (25 mM each)||0.5 ul||1 mM|
|5X Herculase II reaction buffer||10 ul||1x|
|10 uM each of target-specific Surveyor forward and reverse primers (see Experimental Design)||1 ul||200 nM|
|Herculase II Fusion DNA polymerase||0.5 ul||1x|
|Distilled water||37.5 ul|
CRITICAL STEP Surveyor procedure (Steps 46Aiii–xv) is carried out according to the manufacturer’s protocol and is described in greater detail in the Surveyor manual. We provide brief details here since mutation detection by mismatch endonuclease is not a common procedure for most laboratories.
CRITICAL STEP When performing the Surveyor assay for the first time, we suggest carrying out the positive control reaction included with the Surveyor nuclease kit.
|1||95°C, 3 min|
|2–36||95°C, 30 s||55°C, 15 s||72°C, 30s|
|37||72°C, 5 min|
CRITICAL STEP If multiple amplicons are generated from the PCR reaction, re-design primers and re-optimize the PCR conditions to avoid off-target amplification.
|1||95°C, 10 min|
|2||95°C to 85°C, −2°C/s|
|3||85°C, 1 min|
|4||85°C to 75°C, −0.3°C/s|
|5||75°C, 1 min|
|6||75°C to 65°C, −0.3°C/s|
|7||65°C, 1 min|
|8||65°C to 55°C, −0.3°C/s|
|9||55°C, 1 min|
|10||55°C to 45°C, −0.3°C/s|
|11||45°C, 1 min|
|12||45°C to 35°C, −0.3°C/s|
|13||35°C, 1 min|
|14||35°C to 25°C, −0.3°C/s|
|15||25°C, 1 min|
|0.15 M MgCl2 solution||2 ul||15 mM|
|Surveyor Nuclease S||1 ul||1x|
|Surveyor Enhancer S||1 ul
|Re-annealed duplexes from Step 46Aviii||16 ul|
PAUSE POINT: The digestion product can be stored at −20°C for analysis at a later time.
This calculation can be derived from the binomial probability distribution given a few conditions: that strand reassortment during the duplex formation is random, that there is a negligible probability of the identical mutations reannealing during duplex formation, and that the Surveyor nuclease digestion is complete.
CRITICAL STEP Use proper RNA handling techniques to prevent RNA degradation, including cleaning bench surfaces and pipettes with RNAseZAP. Use RNAse-free consumables and reagents.
CRITICAL STEP Do not leave the cells in trypsin for longer than a few minutes.
CRITICAL STEP Incomplete removal of the supernatant can result in inhibition of cell lysis. PAUSE POINT: Cells can be frozen at −80°C for 24 hours.
CRITICAL STEP Protect the probes from light and do not allow the thawed probes to stay on ice for an extended time.
|Cycle number||Denature||Anneal and Extend|
|1||95°C, 20 s|
|2–41||95°C, 1 s||60°C, 20 s|
CRITICAL STEP The ΔΔCT method assumes that amplification efficiency is 100% (ie. number of amplicons doubles after each cycle). For new probes (such as custom TaqMan probes), amplification from a template dilution series (spanning at least 5 orders of magnitude) should be performed to characterize amplification efficiency. For standard TaqMan Gene Expression Assay probes, this is not necessary as they are designed to have 100±10% amplification efficiency.
Troubleshooting advice can be found in Table 3.
|4||Uneven amplification across monomers||Not using Herculase 2 Fusion polymerase||Optimize annealing temperature and Mg2+ and DMSO concentrations|
|8||Low DNA concentration after elution||Residual ethanol on purification column||Air dry columns before elution at 37°C for a longer period of time|
|Incorrect vacuum pressure during DNA binding||Adjust vacuum pressure according to the manufacturer’s suggestions|
|15||No visible hexamer band (~700 bp)||Not adding equimolar amounts of monomers||Gel normalize monomer concentration|
|Degraded DTT or ATP||Use fresh stocks of DTT and ATP, which degrade easily|
|No visible hexamer band (~700 bp) but smaller bands present||Wrong monomer(s) added during pipetting||Re-select monomers|
|Monomer concentration is too low||Increase the number of Golden Gate digestion-ligation cycles and/or increase the concentration of monomers to >20 ng/ul; there is no detrimental effect to using more monomers in an equimolar ratio|
|20||No visible hexamer band (~700 bp)||Unsuccessful Golden Gate digestion-ligation||Verify on a gel that the Golden Gate digestion-ligation product from Step 15 is visible; increase monomer concentration|
|24||Low concentration for purified hexamers||Unsuccessful gel extraction||Ensure that there is no residual ethanol during elution or increase PCR reaction volume|
|28||No visible 18mer band (~1.8 kbp)||Unsuccessful Golden Gate digestion-ligation||Increase hexamer concentration in Golden Gate digestion-ligation in Step 26 or proceed directly to transformation in Step 29|
|30||More than a few colonies on negative control plate||Compromised TALE backbone||Perform a restriction digest of the backbone to verify integrity|
|35||Colony PCR bands are smeared||Too much template||Dilute colony suspension 10x–100x|
|38||Monomers assembled in incorrect order||Misligation||Misligation occurs at a very low frequency; analyze two additional clones|
|45||Low transfection efficiency||Low DNA quality||Prepare DNA using high quality plasmid preparation|
|Suboptimal DNA to Lipofectamine2000 ratio||Titrate DNA to Lipofectamine2000 ratio to determine optimal transfection condition|
|46Av||Multiple amplicons||Nonspecific primers||Design new primers and verify specificity using PrimerBLAST; use touchdown PCR|
|No amplification||Suboptimal PCR condition||Optimize annealing temperature and Mg2+ and DMSO concentrations|
|46Axiii||No cleavage bands visible||TALEN unable to cleave target site||Design new TALEN pairs targeting nearby sequences|
|46Bxii||No increase in transcription in target mRNA||TALE-TF unable to access target site||Design new TALE-TFs targeting nearby sequences|
TALE-TFs and TALENs can facilitate site-specific transcriptional modulation3–5, 8 and genome editing4, 7, 9, 11–15 (Table 1). TALENs can be readily designed to introduce double-stranded breaks at specific genomic loci with high efficiency. In our experience, a pair of TALENs designed to target the human AAVS1 locus is able to achieve up to 3.6% cutting efficiency in 293FT cells as determined by Surveyor nuclease assay (Fig. 6a–b). TALE-TFs can also robustly increase the mRNA levels of endogenous genes. For example, a TALE-TF designed to target the proximal promoter region of SOX2 in human cells is able to elevate the level of endogenous SOX2 gene expression by up to 5 fold3 (Fig. 6c). The ability for TALE-TFs and TALENs to act at endogenous genomic loci is dependent on the chromatin state as well as yet-to-be-determined mechanisms regulating TALE DNA binding56, 57. For these reasons we typically build several TALE-TFs or TALEN pairs for each genomic locus we aim to target. These TALE-TFs and TALENs are designed to bind neighboring regions around a specific target site since some binding sites might be more accessible than others. The reason why some TALEs exhibit significantly lower levels of activity remains unknown, though it is likely due to position- or cell state-specific epigenetic modifications preventing access to the binding site. Due to differences in epigenetic states between different cells, it is possible that TALEs that fail to work in a particular cell type might work in a different cell type.
Supplementary Data 1. Nucleotide sequences of the 4 monomer plasmids, 4 TALE-TF cloning backbone plasmids, and 4 TALEN cloning backbone plasmids.
We thank the entire Zhang laboratory for their support. L.C. is supported by a HHMI International Student Research Fellowship. Y.Z. is supported by a Simons Foundation Fellowship. M.M.C. is supported by a MIT Undergraduate Research Opportunities scholarship. F.Z. is supported by a NIH Transformative R01, the McKnight and Simons Foundations, Robert Metcalfe, and Michael Boylan.
Author contributionsN.E.S., L.C., Y.Z., and F.Z. wrote the manuscript. M.M.C. designed the online TALE sequence verification software. F.Z. and G.F. supervised the research.
Competing financial interest
The authors declare that they have no competing financial interests.