Peptidomimetic Foldamers

Recently, Gellman and co-workers have developed an (a/b + a)-peptide ligand for the same BH3-recognition cleft of Bcl-xL [62]. For their design, they studied the different helices adopted by oligomers with a 1:1 alternation of a- and b-amino acid residues along the foldamer backbone (see Chapter 2) [63]. Pure b-peptides, as well as a/b-peptide designs, which formed 11-helices promoted by a five-membered ring constrained b-amino acid, showed no activity in FP assays. On the other hand, those a/b-peptides that adopted a 14/15-helix showed some activity. To gain affinity, Gellman et al. replaced either the N or C terminus of their best a/b-peptide inhibitor with a fragment of the natural a-peptide sequence of Bak. When the N terminus was modified no activity was observed. On the other hand, the C-terminal modified chimeric peptide 19 showed a Ki value of 1.9 nM (Fig. 7.14). In this (a/b + a) peptide, a Leu was introduced in the 6th position and a Phe at the 12th position of the 14/15 helix in order to mimic Leu78 and Ile84 of the Bak peptide respectively. In addition, an Arg at position 4 and an Asp at position 11, were included to increase binding via electrostatic interactions with residues on the edge of the Bcl-xL cleft. Because of the proteolytic susceptibility of the a-segment in 19, future work will be aimed at replacing it with a non-natural segment.

The hDM2/p53 complex is another well characterized protein-protein interaction involved in apoptosis [64]. In unstressed normal cells, p53 is present at very

Fig. 7.14 Structure of the (a/b + a) chimeric scaffold.

low levels due to rapid degradation via a ubiquitin-dependent proteasome pathway [65]. In cancerous cells, however, hDM2 is overexpressed abrogating the ability of p53 to induce cell cycle arrest and apoptosis by binding p53 and promoting its degradation [66]. For the inhibition of the hDM2/p53 interaction, Schepartz et al. designed a class of 14-helical b-peptide inhibitors [67, 68]. They sought to mimic three residues on the activation domain of p53 (Phe19, Trp23, and Leu26) which had been shown to be important for the heterodimerization of p53 and hDM2 [69]. In general short b-peptides lack a well-defined secondary structure in water. However, Schepartz et al. were able to obtain well-folded 14-helix structures in water by neutralizing the helix macrodipole. To do this, on one of the three faces of the 14-helix b-peptide scaffold, they introduced a positively charged side chain (b3-homoornithine) near the N terminus and a negatively charged residue (b 3-homoglutamate) at the C terminus while still maintaining a stabilizing salt bridge network (Fig. 7.15). This was the first example where this strategy was used to stabilize short b-peptide 14-helix structures. The second face of the 14-helix was primarily substituted with b3-homovaline residues, while the third face was used to present the functional epitope necessary for protein recognition. FP assays showed that 20 bound to hDM2 with nanomolar affinity and displaced a p53 peptide derivative from hDM2 with an IC50 = 94.5 mM. Scrambling of the b3-homophenylalanine, b3-homotryptophan, and b3-homoleucine resulted in loss of affinity, emphasizing the importance of mimicking the structural features of the side chains in maintaining specificity and affinity.

Fig. 7.15 Structure of 14-helix b-peptide 20 (Reproduced with permission from the authors).

Fig. 7.16 (a) General structure of a peptoid; (b) Peptoid monomer side chains used by Barron and co-workers.

Oligomers of N-substituted glycines, known as peptoids, have also been developed to mimic proteins [51, 70]. In peptoids, the peptidic side chains are shifted to main chain nitrogen atoms that eliminate stereogenic centers and the possibility of forming hydrogen-bonding networks (Fig. 7.16a). Even though intramolecular hydrogen bonding is not possible, the inclusion of appropriate bulky chiral substituents can induce peptoids to acquire secondary structures (see Chapter 1, Section 1.4.1) [71, 72]. In 2005, Barron et al. reported a class of helical peptoids that mimic the lung surfactant (LS) protein B (SP-B) [73]. During respiratory distress, SP-B interacts with the lipid film by an unknown mechanism to enable breathing. SP-B is an amphipathic 79-amino acid protein with four disulfide bonds. Three of the disulfide linkages are intramolecular, while the fourth is intermolecular forming a homodimer. The presence of multiple disulfide bonds in the SP-B homodimer complicates its synthesis, thus synthetic mimics of SP-B are currently being investigated for use as additives or biomimetic lung surfactant replacements.

In an effort to synthesize a peptoid mimetic of SP-B, Barron and co-workers modeled their design after a small amphipathic N-terminus segment of SP-B (SP-B1 25) that had been shown to retain much of the surfactant function. They designed three 17-mer peptoids based on earlier reports describing the structural properties of peptoids whose sequence incorporated certain bulky a-chiral aromatic and aliphatic side chains as shown in Fig. 7.16b. The first peptoid included exclusively a-chiral aromatic side chains with an achiral lysine-like monomer at every third position to create a cationic face and reproduce the amphipathicity of SP-B1a25. The second incorporated both a-chiral aromatic and aliphatic residues in addition to the lysine-like monomers. Finally the third mimetic consisted of only a-chiral aliphatic side chains as well as the lysine-like monomers. As expected from earlier reports, the CD spectra of the first two peptoids in water showed features corresponding to a poly-proline type I-like helix. The third was found to be a random coil due to the lack of a-chiral aromatic monomers though it is the most similar to SP-B1 25 in that it contains no aromatic residues. All three peptoids, however, exhibited spectra characteristic of helices in a lipid environment.

Four different in vitro techniques were used to characterize the surface-active properties of the peptoids. Their activities were compared to two peptides (SP-B1 25 and KL4) that had been previously established, by in vivo as well as in vitro studies, to be good mimics of SP-B. Extensive in vitro characterization of the pep-

toids' surface activities showed that their properties were similar to those previously reported for other SP-B mimetics. Barron and co-workers also reported that this similarity appeared to correlate with the overall helicity and hydrophobic-ity of a given peptoid.

The scope of peptidomimetic compounds is not limited to mimicking a-helices. In fact, b- and /-peptides have also been used to mimic b-turns (see Chapter 2, Section 2.5). One of the earliest examples where a b-peptide was used to recognize a biomacromolecule was reported by Seebach and co-workers [74]. They developed a cyclic b-tetrapeptide analog of octreotide that would bind to human somatostatin receptors. Octreotide is a cyclic a-octapeptide derived from somatostatin that is currently in use to treat acromegaly and intestinal cancers [75, 76]. SAR studies performed on octreotide revealed that the amino acids in the b-turn (Phe-Trp-Lys-Thr) were required for activity [76]. Though the affinity of the cyclic b-tetrapeptide was an order of magnitude lower than octreotide, this study demonstrated that a b-peptide could be used as a b-turn mimetic.

More recently, Seebach and co-workers designed an unconstrained b-peptide that would be predisposed to fold [77]. They incorporated an a- branched (b 2)/b-branched (b3) b-dipeptide sequence into their design which had been shown to induce a b-turn structure with a 10-membered intramolecular hydrogen bond [78]. They made a tetrapeptide 21 (Fig. 7.17) as a potential somatostatin (SRIF) mimic by attaching a Lys and a Trp residue to the b2 /b3 motif, while the other two residues were b3-HPhe and b3-HThr. Two other peptides were synthesized for comparison purposes. The first was an all-b3-amino acid peptide lacking the b2/b3 motif 22 and it was expected to fold into an a-helix. The second was an all-a-amino acid peptide expected to be unstructured in solution. One-dimensional NMR spectroscopy and CD were used to obtain information about the secondary structures of the peptides. In the JH NMR, the C(g)-H protons of peptide 21 are shifted upfield in agreement with a b-turn conformation where its C(g)-H protons are in close proximity to the tryptophan indole ring. The C(g)-H protons of b3-peptide 22, showed no such upfield shift indicating the absence of a b-turn. In the CD spectra of 22 and the a-peptide no Cotton effect was observed, while 21 shows a maximum at 203 nm in methanol. This agrees well with the spectrum of a well-characterized b-peptide turn.

Fig. 7.17 Structure ofb-peptides 21 and 22.



Fig. 7.18 (a) Structure of one of the four g-peptides 23; (b) Proposed conformation of 23 showing b-turn structure.

Fig. 7.18 (a) Structure of one of the four g-peptides 23; (b) Proposed conformation of 23 showing b-turn structure.

A radioligand-binding assay was performed to evaluate the affinity of the peptides for five human recombinant SRIF receptors. Both the linear b3-peptide 22 and the a-peptide showed low affinity for the receptors with Kd values higher than 10 mM. On the other hand, 21 showed a significant and selective affinity for the human receptor sst4 (Kd = 83 nM). Octreotide was found to have 20 times higher affinity than 21, whereas somatostatin had an affinity 20 times lower. Interestingly going from peptide 21 to 22 ,which only differ in the position of the Lys chain, resulted in nearly a 1000-fold decrease in the affinity for sst4.

Seebach and co-workers have developed a series of g-dipeptide derivatives that also target human somatostatin receptors [79]. They designed four peptides that included a tryptophan sidechain in the g2 position of the first g-amino acid and a lysine in the g4 position of the second (Fig. 7.18). Similar to the b-peptides previously discussed, the NMR of all four g-dipeptides prepared confirmed the proximity of the lysine chain and tryptophan ring, indicative of a turn conformation as seen for 23 in Fig. 7.18b. In addition, the CD spectra of the peptides exhibited an intensive negative Cotton effect further supporting the idea of the presence of secondary structure. The peptides were tested in a radioligand-binding assay against five human somatostatin receptors. Their best inhibitor exhibited a Kd value of 510 nM.

Synthetic foldamers can also be used to mimic the quaternary structural elements of proteins. In this approach, the recognition occurs between two or more synthetic partners yielding an all-artificial structure that resembles a protein-like assembly. Recently, Schepartz et al. have reported the recognition between structured b-peptide helices that form a complex [80]. Two b-peptide monomers, Acid-1F and Base-1F, were designed to recognize each other (Fig. 7.19). Both contained free N and C termini as well as one face that included alternating positively and negatively charged residues to stabilize the 14-helix conformation. b 3-homoleucine residues were incorporated into positions i and i + 3 of the 14-helix and b3-homophenylalanine residues were incorporated at positions 4 and 7 to favor interhelical interactions. In addition, at positions 1 and 10, a

0 0

Post a comment