Saturday, 22 December 2012

Artificial transcription factors

A General strategy for selecting high-affinity zinc finger proteins for diverse DNA target sites by Greisman and Pabo
Gradually assemble new protein at desired binding site. Add and optimise one finger at a time as they proceed to target site.  Use Zif268 structure.

3 selection steps are used. One for each finger of new protein
1) A finger that recognises 3'end of target site is selected by phage display.  2 wt Zif fingers are used as temporary anchors to position library of randomised fingers over target site. Use hybrid DNA site with Zif subsites fused to target site
2) Selected finger is retained as part of growing protein. After distal Zif finger is discarded, phage display selects a new finger that recognises central region of target site.
3) Remaining Zif finger is discarded. Phage display selects a 3rd finger. It recogninses 5' region of target site. Optimisation yields new zn finger protein.

New fingers are selected in relevant structural context. An intact binding site is present at every stage. Selections are performed in context of growing protein-DNA complex.  Selects for fingers that function well together. To ensure that selection proteins will bind tightly to desired target sites, do all seelctions in presence of calf thymus competitor DNA. This counterselects against proteins that bind promiscuously or prefer alternative sites.
Fig. 2.
Overview of protocol that successively selects finger 1, finger 2, and finger 3 to create a new zinc finger protein. Fingers that are present in the phage libraries used in these steps (15) are indicated on the left side of each panel. Zif1 and Zif2, wild-type Zif268 fingers; R, a randomized finger library; and asterisk, a selected finger. Small horizontal arrows indicate the multiple cycles of selection and amplification used when selecting each finger by phage display (35). The right side of each panel shows the binding sites used in selections with the TATA site and indicates the overall binding mode for the selected fingers [each DNA duplex has biotin (not shown) attached at the 3′ end of the upper strand]. Vertical arrows indicate how fingers selected in earlier steps are incorporated into the phage libraries used in later steps and reselected to optimize affinity and specificity in the new context (16). (A) A randomized finger 1 library was cloned into the pZif12 phagemid display vector (36), and selections with this library were performed in parallel at the TATA, p53, and NRE sites (17). (B) The wild-type Zif1 finger was removed, and a randomized finger 2 cassette was ligated to the appropriate vector pool and optimized by phage display (29). (C) The remaining wild-type finger was removed, and a randomized finger 3 cassette was added and optimized by phage display. To construct the sites used in these selections, we fused the target strand with the higher purine content to the guanine-rich strand of the Zif268 site. Because of the overlapping base contacts that can occur at the junction of neighboring subsites (Fig. 1B), the 3′ end of the target site (Fig. 1C) was aligned so that it overlapped with the Zif2 subsite.
Selections were performed with a TATA box, p53 binding site and a nuclear receptor element (NRE). These regulatory sites are normally recognised by other families of DNA-binding proteins. The sites are different from guanine-rich Zif268 site.

Fig. 3.
Amino acid sequences of new zinc finger proteins that recognize (A) the TATA box, (B) the p53 binding site, and (C) the NRE. Residues selected at each of the six randomized positions are shown (37). Six or more of the eight clones in each phage pool encode unique zinc finger proteins (1619). A box indicates the clone that was overexpressed and used for binding studies. Residues that are fully conserved (eight of eight clones) are shown in boldface; residues that are partially conserved (four or more of eight) are denoted by lowercase letters in the consensus sequence below the set of clones. Modeling (38) suggests that these new zinc finger proteins (including those that recognize the TATA box) can bind to B-form DNA. Each panel indicates how the fingers could dock with a canonical 3-bp spacing (dashed boxes), and dashed arrows indicate plausible base contacts (2026). Recent data from studies of a designed zinc finger protein provide precedence for many of these contacts (39). Detailed modeling suggests many additional contacts (not shown), including some that couple neighboring fingers and subsites (38). For the p53 site, there is an alternative, equally plausible, docking arrangement with a 4-bp spacing for one of the fingers (40). A section of the NRE site shows a 5 of 6 bp match (underlined) with the Tramtrack binding site, and these matching segments happen to be aligned such that the new fingers bind in the same register as the Tramtrack fingers (11). Every Tramtrack residue that contacts one of the matching bases (solid arrows) was recovered in our selections (26). Two residues that do not directly contact the DNA in the Tramtrack complex were also recovered (at positions 5 and 6 in NRE finger 3).

Modular design of articifial transcirpiton factors by Ansari and Mapp
Modular nature of eukaryotic TFs
Natural transcriptional regulators have DNA binding domain and regulatory domain (DBD and RD). DBD causes pseicificty in targeting RD to spec site in genome. REgulatory mediate effects on gene where theyt are delivered.

DBD are characterised by structure (HTH) but RDs are catgegorised by abundance of amino acids eg acid rich, gluatmin-erich, prioline rich etc. RDs are characterised by functional context. Some activator and repressor modules are categorised according to distance from promoter where they funciton.

Modular design of ATFs
DNA binding domains
Protein DBDs are attractive targets as they have high specificity and affinity for  target sequences. No clear recognition code has been discoverd. Use genetic selection to generate protein DNa binding modules for spec sequence. Zn finger DBDs that target unique promoters have been isoalted and function in cultured cell lines.

Regulatory domain: activation
Activatin modules with multiple acidic and hydrophobic residues function robustly in all eukaryotes tested. Often used in ATAFs. Designed zinc fingers fused to acidic activating domain eg VP16, a potent viral coactivator upregulates genes.

Regulatory domains: repression
Fuse large segments of repressor proteins to DBD. eg KRAB repressor moduls us fused to designed zinc fingers, a modest repression of adjacent gene.

Inhibit by competition for DNA binding site of an endogenous TF. Polyamides which targeted TF binding sites inhibited expression of HIV genes in cell culture.

Linking DNA binding and regulatory domains
For active activation or repression DBD and RD must be linked together to function as an ATF. Tether with peptide linekrs.

Traffic and delivery
Obstacle. Trafficking of polyamide ATFDs into cell nucleir depends on cell type/ TFO-bsed regulators are devliered by electroporation or cationic liposomes. Unsusatinable and limits tissues it can be targeted. Attaching NLS is effective.

Fig. 3. Modules used in the design of ATFs. Representative examples of the DBDs, RDs and linkers used in the design of most ATFs are summarized. Asterisks denote modules for which little structural data are available. Common protein-based DBDs include Gal4 [59], helix-turn-helix proteins (HTH; a generic binding scheme is shown) such as LexA, LacI and TetR, and both natural and designed zinc-finger motifs (structure from [60][11••]. The most frequently used non-protein DBDs are TFOs, PNAs and polyamides (for structure see [61]). While the linker is most commonly a flexible (GGSGGS) or rigid (polyproline [50]) peptide, a simple flexible linker derived from polyethylene glycol (PEG) can also be used without loss of function. Small molecules such as the FK506-CsA dimer [49] or the dimer of methotrexate (Mtx) and a synthetic analog of FK506 (SLF) [52] can also bridge the DBDs and RDs by non-covalent interactions [48]. Most activating regions have been derived from natural activator proteins such as VP16  and  or designed to mimic natural activating regions (AH)  and . One exception is the use of RNA hairpins [39] (S Saha, AZ Ansari, K Jarrell and M Ptashne unpublished results). The repression domain most often used is derived from the KRAB protein  and  but small peptides such as WRPW [40] and CCVC [41] can also function as repressors when bound to DNA. In many cases, the modules shown can be readily swapped to generate ATFs that target different DNA sequences or have a different regulatory function.

No comments:

Post a Comment