|Home | About | Journals | Submit | Contact Us | Français|
Engineered designer nucleases can be used to efficiently modify genomic sequence in a wide variety of model organisms and cell types. Zinc finger nucleases (ZFNs), consisting of an engineered zinc finger array fused to a non-specific cleavage domain, have been extensively used to modify a broad range of endogenous genes. Here we describe protocols for engineering ZFNs targeted to specific gene sequences of interest using the Context-Dependent Assembly (CoDA) method.
Engineered zinc finger nucleases (ZFNs) can be used to introduce site-specific genome alterations in a wide variety of model organisms and a range of different mammalian cell types (Rahman et al., 2011; Urnov et al., 2010). ZFNs are customizable restriction enzymes that can be used to create targeted double-stranded breaks (DSBs) in cells. ZFNs consist of an array of engineered zinc fingers fused to the non-specific cleavage domain of the FokI Type IIS restriction enzyme (Kim et al., 1996). The engineered zinc finger portion of the ZFN directs the nuclease activity to a specific location in the genome. Because the FokI domain must dimerize to cleave DNA, a pair of ZFNs must be engineered to cleave a given target site of interest. Each monomer in a ZFN dimer pair binds to a “half-site” with cleavage of the DNA induced in an intervening 5–7 bp “spacer” sequence.
The repair of ZFN-induced DSBs by normal mechanisms used by nearly all cells can be exploited to make targeted genomic alterations. For example, non-homologous end-joining mediated repair of a ZFN-induced break can lead to the efficient introduction of insertion or deletion mutants, providing a means to create frameshift and knockout mutations if introduced into coding sequences of endogenous genes (Bibikova et al., 2002; Santiago et al., 2008). In addition, ZFN-induced DSBs can also be repaired by homologous recombination with an investigator-supplied “donor template” (Bibikova et al., 2003; Urnov et al., 2005). This template consists of DNA sequences homologous to endogenous sequence surrounding the DSB but with mutations or insertions of interest also incorporated. Repair of the ZFN-induced DSB with this donor template can thus be used to introduce particular mutations or insertions into specific genomic loci.
Three methods have been described (Beerli and Barbas, 2002; Maeder et al., 2008; Sander et al., 2011) for engineering zinc finger arrays required to make ZFNs (Figure 1). These methods are known as (1) modular assembly, (2) oligomerized pool engineering (OPEN), and (3) context-dependent assembly (CoDA). Each of these methods possesses different advantages and disadvantages that we discuss briefly here.
With modular assembly (Figure 1C), individual pre-selected fingers are joined together to create a zinc finger array (Beerli and Barbas, 2002). This method relies on archives of pre-selected individual zinc fingers and assumes that the behaviors of fingers in an array are independent of one another. The advantage of this method is the ease with which it can be practiced - arrays can be engineered in as little as a week. However, the failure rate of this method for constructing pairs of three-finger ZFNs has been reported to be very high (>94%) (Kim et al., 2009; Ramirez et al., 2008). In addition, three-finger arrays produced by modular assembly have also been shown to have low affinities, low specificities, and low activities as ZFNs in cells (Cornu et al., 2008; Hurt et al., 2003; Ramirez et al., 2008). Various detailed protocols for practicing modular assembly have been previously described (Carroll et al., 2006; Gonzalez et al., 2010; Wright et al., 2006).
The OPEN method, unlike modular assembly, explicitly accounts for the context-dependent activities of individual zinc fingers in a three-finger array (Maeder et al., 2008; Maeder et al., 2009). The method relies on the creation of a combinatorial library of multi-finger arrays derived from “pools” of pre-selected fingers for individual three bp “subsites”. Three pools are randomly recombined together to create a library and then a bacterial two-hybrid selection system is used to identify specific library members that bind to a target site of interest (Figure 1A). The OPEN method is highly efficient and has yielded ZFNs that function efficiently in zebrafish, plants, and human somatic and pluripotent stem cells (Foley et al., 2009; Maeder et al., 2008; Townsend et al., 2009; Zhang et al., 2010; Zou et al., 2009). Although OPEN represents a substantially simpler protocol to practice over previously described selection-based methods, it remains lengthy and challenging for most laboratories to practice. The method takes approximately 2 months to complete and requires a substantial investment of time (typically 6 to 12 months) to successfully master. A detailed protocol for practicing OPEN has been previously described (Maeder et al., 2009).
The CoDA method is an assembly-based method that attempts to account for the context-dependent activities of adjacent zinc fingers in an array (Sander et al., 2011). With CoDA, three-finger arrays are engineered by assembling together fingers that have been pre-selected to function well together. As shown in Figure 1B, an amino-terminal finger (F1) is joined to a particular middle finger (F2) with which it is known to work well. A third carboxy-terminal finger (F3) is then added that is known to work well with the same F2. The resulting three-finger arrays are then screened for DNA-binding activity using a bacterial two-hybrid assay. This method is as simple to practice as modular assembly but is also highly efficient with approximately 50% of ZFN pairs showing activities in zebrafish and plants (Curtin et al., 2011; Sander et al., 2011).
Researchers interested in engineering their own ZFNs face the initial challenge of choosing from among the three methods described above. In general, we recommend the use of CoDA as a first-line approach. CoDA is simple to practice and has a high success rate for yielding functional ZFNs. CoDA ZFNs have been used successfully in zebrafish and plants and unpublished data from our lab shows that these nucleases also function in transformed human cell lines. The one exception to our recommendation of CoDA as the initial method of choice is for those investigators requiring ZFNs with the highest possible activities and specificities. Although CoDA does account for context-dependent effects between adjacent fingers, in our experience the OPEN method generally yields ZFNs with higher activities. Presumably OPEN ZFNs also possess higher specificities than those made by CoDA because the selection process interrogates a larger number of combinations of fingers and identifies those arrays that can find their target site in the context of competing E. coli genomic DNA.
The CoDA method will not yield ZFNs for every target gene of interest. CoDA currently has a targeting range of approximately 1 in every 500 bps of random DNA sequence. If one is unable to identify ZFNs using CoDA, we recommend that users attempt to use OPEN, which has a targeting range of approximately 1 in every 200 bps of random DNA sequence. If no potential OPEN ZFN target sites can be identified or if OPEN selections fail to yield ZFNs, users can consider using modular assembly with the caveat that the success rate may be low.
This section describes in detail how to practice the Context Dependent Assembly (CoDA) for engineering zinc finger proteins. CoDA reagents are currently capable of targeting one site in every 500 bp of random sequence, a range sufficient for generating non-homologous end joining (NHEJ) frame shift mutations for most genes in many organisms. CoDA ZFs are designed using zinc finger domains that have been selected to work well together. The resulting DNA sequences encoding CoDA ZF arrays are less than 300 bp in length and can be synthesized and subsequently cloned into standardized expression plasmids for testing binding affinity in the bacterial two-hybrid (B2H) system or for use as site-specific nucleases.
Repeat elements will not be identified if the submitted sequence is shorter than the repeat element. To ensure that the repeat elements are identified submit the genomic sequence that includes several hundred base pairs on both sides of your region of interest.
Use genomic sequence (not CDS) to ensure the ZFN targets are present in adjacent sequence.
It is also recommended to use actual DNA sequence obtained from the host organism to ensure there are no polymorphic variations from the published sequence.
By default, ZiFiT defines any sequences in uppercase as exons and sequences in lowercase as introns. This can be used later to aid in picking targets. This option can be deactivated by unchecking the box labeled “Exon/Intron Case Sensitivity” below the lower right corner of the sequence box.
Alternatively, select targets by clicking the blue, green, and gold bars representing targets color coded by spacer size (5bp=Blue; 6bp=Green; 7bp=Gold) above the red bars (thin bar = introns; thick bar = exon) from the graphic summary pop-up window. Pop-ups must be enabled in your web browser for this feature to work. Use of the back button can disrupt the link between the popup and the main window. Restore functionality by closing the popup and selecting the ZiFiT link from the menu.
Use the following criteria to maximize success rates 1) Choose target sites where each of the two half-sites are composed of two or three GNN subsites 2) Avoid targets with half-sites that are composed of four or more T’s 3) Click on the color-coded array in the double-strand DNA target sequence to query ZiFDB and determine whether a zinc finger array for a half-site has already been assayed for DNA-binding activity. Use arrays that have been previously shown to activate three-fold or more in the bacterial two-hybrid assay. 4) Pick targets early in coding sequence to maximize likelihood of generating a knockout mutation. 5) Pick target sites within 100bp of the location of the alteration to be introduced by homologous recombination. 6) Pick two or three ZFN target sites per gene of interest.
To confirm that a zinc finger array can bind to its target site, zinc finger arrays are expressed in B2H reporter strains and assayed for their abilities to activate expression of a beta-galactosidase reporter gene. The B2H reporter strain harbors a single-copy plasmid with the zinc finger array target site positioned upstream of a weak promoter which, in turn, controls expression of a lacZ reporter gene (encoding beta-galactosidase). Binding of the zinc finger array (fused to a Gal11P protein fragment) to its target site will recruit RNA polymerase to the promoter, leading to an increase in expression of beta-galactosidase (Figure 2). To perform this experiment, a B2H reporter strain is transformed with two plasmids; one expressing a zinc finger array-Gal11P fusion protein and one expressing an RNA polymerase-alpha-Gal4 hybrid protein. Because the lacZ gene encodes β-galactosidase, its expression can be measured by a simple quantitative assay. By comparing β-galactosidase expression in the presence and absence of a given zinc finger array, a “fold-activation” value can be calculated which can guide the choice of which arrays to carry forward for testing as ZFNs.
If your zinc finger proteins contain a NotI site instead of a BamHI site for cloning into the vectors pMLM800 or pMLM802, you will need to first amplify them by PCR (CPMB UNIT 15.1) to generate a BamHI site using the following primers:
- Forward: 5’- GAAAAAAATCTAGACCCGGGGAGC -3’
- Reverse: 5’- CGCGGATCCCCTCAGGTGGGTTTTTAGGTG-3’
Subculturing the cells containing pBAC-LacZ plasmids in arabinose induces expression of the trfA gene product, which increases the copy number of the plasmid.
|1 cycle:||5 minutes||95°C||(initial denaturation)|
|35 cycles:||30 seconds||95°C||(denaturation)|
|1 cycle:||5 minutes||72°C||(final extension)|
Shaking on a platform with a small throw radius of no more than a few millimeters (for example, Microtitertron orbital shaker, Appropriate Technical Resources) will ensure uniform growth of all wells. Overnight culture time should not exceed 18 hours.
For best results lyse as close to and OD600 of 0.3 as possible.
The activity of β-galactosidase is stable in the cell lysates for at least 18 hours when stored sealed at room temperature.
Many microtiter plate readers can be programmed to take absorbance measurements at fixed intervals and to calculate the velocity of ONPG cleavage. We typically take measurements every 10–30 seconds. Reactions should not be allowed to proceed for more than 30 minutes as substrate can become limiting.
|V × (1000) /|
In this step, each ZF array is cloned into a nuclease expression vector. Expression vectors differ based on the type of promoter, nuclease domain, and linker between the ZF and nuclease domain (see Table below). The linker determines the number of nucleotides (5, 6, or 7) that are permitted in the “spacer” sequence between the ZF target half-sites (Handel et al., 2009). The wild-type FokI nuclease monomer permits ZF monomers from the same half-site to form homodimers. This increases the likelihood that the ZFNs will cleave at unintended positions. Alternatively, the heterodimer variants require two different ZFNs to assemble on a full target site to mediate cleavage (Miller et al., 2007). However, the heterodimer variants used in the expression plasmids described below are less active than the wild-type FokI domains.
|Plasmid name||For expression in||FokI cleavage|
|Wild-type||5 or 6|
|5 or 6|
|5 or 6|
|pDW1775||Plants||Wild-type||5 or 6|
Sander, J.D., Dahlborg, E.J., Goodwin, M.J., Cade, L., Zhang, F., Cifuentes, D., Curtin, S.J., Blackburn, J.S., Thibodeau-Beganny, S., Qi, Y., et al. (2011). Selection-free zinc-finger-nuclease engineering by context-dependent assembly (CoDA). Nat Methods 8, 67–69.
Internet Resources: ZiFiT (http://zifit.partners.org)