Design Of Preorganized Structures

Although the entropic cost of folding a single chain is less than that required for the intermolecular association of modules, the cost of forming a unique three-dimensional structure from an unconstrained linear chain is still substantial. The unfavorable con-formational entropy change associated with folding into a unique structure can be decreased by preorganizing the desired structure. Several groups have employed this strategy by using covalent cross-links, synthetic templates, or ligand binding.

Linking peptides to a synthetic template is a non-natural but effective way to covalently constrain peptides to adopt a particular structure. Mutter and coworkers pioneered this approach by developing an artificial template suitable for the formation of template-assembled synthetic proteins (TASPs; see Figure 11.1C) (Mutter and Vuilleumier 1989; Altmann and Mutter 1990). Their template was a short oligopeptide designed to form two antiparallel P-strands, stabilized by a disulfide bond at the open end. The e-amino groups of four lysine side chains located on the template served as attachment sites for four units of secondary structure. Thus, four a-helices could be attached to form a parallel four-helix bundle with considerable stability. Ultimately, Mutter and coworkers were able to construct a variety of TASPs with different stoichiometries and secondary structures, including pap and 4a/4p structures (Mutter et al. 1989). Using similar strategies, Haehnel and coworkers have designed a series of template-based structures, including several with biologically relevant activities (Li et al. 2006).

In such designs, the template does not have to be a peptide. Thus, Sasaki and Kaiser used a porphyrin ring to attach four copies of an amphiphilic peptide (Sasaki and Kaiser 1989; Sasaki and Kaiser 1990). The resulting four-helix bundle heme protein (called Helichrome) was as stable as many natural proteins and also possessed a low level of enzymatic activity.

The introduction of ligand binding sites into secondary structural units can also be used to drive assembly into globular protein structures (Figure 11.1D). Both Lieberman and Sasaki (1991) and Ghadiri et al. (1992) designed a non-natural divalent metal-binding moiety (bipyridine) onto the N-terminus of a short amphiphilic a-helix, thereby inducing self-assembly into a three-helix bundle upon addition of metals such as Fe(II), Ni(II), Co(II), or Ru(II). Ghadiri's group elaborated on this strategy to assemble four-helix bundles by utilizing the high affinity of Ru(II) for nitrogen-containing aromatic heterocyles attached to the termini of four a-helices (Ghadiri et al. 1992b).

0 0

Post a comment