Globally Modified Peptides
Protein interaction domains have evolved to accommodate specific sequences of preferred amino residues on their protein-binding partner. However, each amino acid site on the bound sequence is limited to a genetically encoded molecular diversity of 20. In reality, the latter is slightly larger due to a small array of possible post-translational modifications. Nevertheless, it is not difficult to imagine that there exist a wealth of potential binding interactions that lie just outside of the reach of this limited set of naturally occurring residues. One strategy to enhance molecular diversity is to prepare a wide assortment of Fmoc and side chain-protected, unnatural, amino acid derivatives and then synthesize the corresponding library of peptides. However, a reasonable 50-fold enhancement in molecular diversity vis-à-vis genetically encoded residues would require the preparation of 1,000 different monomers, a nontrivial task.
Lawrence and his colleagues outlined an approach that creates high diversity at any desired site along a peptide chain using readily available carbox-ylic acid-containing compounds (Lee and Lawrence 1999). The strategy, as outlined in Scheme 3, employs a consensus sequence peptide containing 2,3-diaminopropionic acid ("Dap"), appropriately inserted at key sites along the peptide chain.
1. Solid phase peptide synthesis j-NH2
peptide-Dap-peptide- HN(CH2)2SS(CH2)2NH-0 36
2. Dap side chain deprotection
1. Transfer to each well of 96 well filter plates
2. Acylate with ~103 RC02H
Dithiothreitol assay buffer
peptide-Dap-peptide- HN(CH2)2SS(CH2)2NH-Q 37
The peptide is synthesized on a disulfide-substituted Tentagel resin. Once prepared, the peptide-resin is distributed in equal amounts to the individual wells of a 96-multiwell plate designed for organic synthesis (i.e., the bottom of each well contains a frit that allows multiple washings without loss of the peptide-resin). Each well is then exposed to one of approximately 1,000 different carboxylic acid compounds. In short, the library is prepared in parallel, thereby obviating the necessity of molecular deconvolution. Once the modification in the Dap residue is complete, any side chain protecting groups on the peptide are removed with trifluoroacetic acid. After multiple washings to remove residual acid, the individual peptides are cleaved from the resin with assay buffer, which contains dithiothreitol. The peptides can then be directly assessed for potency.
Lawrence and his team employed a structure-based strategy, in combination with the synthetic approach outlined in Scheme 3, to identify high-affinity ligands for the SH2 domain from Lck (Lee and Lawrence 1999, 2000; Yeh et al. 2001). The three-dimensional structure of the Lck SH2 ligand, ace-tyl-pTyr-Glu-Glu-Ile-amide, bound to its protein target, had been previously reported (Tong et al. 1996). Three sites on the ligand, the N-terminal acetyl moiety and the two Glu side chains, are oriented into regions of the SH2 surface that could potentially accommodate modified analogs of the acetyl and glutamic acid moieties.
The initial library and its subsequent screen furnished compound 39, which contains a coumarin moiety in place of the former N-acetyl group. The coumarin-derivatized peptide exhibits a KD of 35 nM for the Lck SH2 domain, approximately two orders of magnitude greater than the parent
peptide aceyl-pTyr-Glu-Glu-Ile-amide. Subsequent identification of the glutamic acid replacements furnished 40, which displays a KD of 200 pM, approximately four orders of magnitude better than the starting consensus peptide. An analogous approach was recently used to construct an inhibitor for the a isoform of PKC (Lee et al. 2004). The inhibitor displays a Ki of 800 pM and a selectivity of greater than 400-fold versus other conventional, novel, and atypical PKC isoforms.
Lawrence, Zhang, and their colleagues reported a variation on the Scheme 3 strategy that provided a high-affinity inhibitor for PTP1B (Shen et al. 2001). Once again, a structure-based approach was employed that directed molecular diversity toward potential binding sites on the target protein surface. In this particular case, PTP1B had been previously shown to bind phosphotyrosine at two distinct sites, one at the active site and the other at a position adjacent to the active site (Puius et al. 1997).
A library of the general form 41 was prepared using the disulfide Tentagel resin 35. Molecular diversity was inserted at the N-terminal and linker positions. The lead compound 42 exhibits a Ki of 2.6 nM for PTP1B and a selectivity of between 1,000- and 10,000-fold versus a panel of fifteen other protein phosphatases.
Was this article helpful?