2. Antimicrobial Peptide Stapling
One focus of investigations into stapled peptides involves improving protein–protein interaction approaches for the development of anticancer and antiviral drug candidates
[27][28][29][30][31][32][33][34][35]. Although stapling has not been widely used for AMP modifications, this technique has shown to be of great value in improving stability, and it is a promising technique for AMP rational design
[36]. Despite a rising number of AMPs described, due to a series of drawbacks, only a few natural or non-natural AMPs available on different databases undergo clinical trials. The application of stapling strategies could change this scenario for the better
[7].
The Data Repository of Antimicrobial Peptides (DRAMP) contains diverse annotated AMPs and an interesting classification method. Among all 22,259 entries, 5891 are general AMPs (natural and synthetic), 16,110 are from deposited patents, 77 are AMPs in drug development and 181 are stapled AMPs (
Figure 2A). Analysis of the 181 stapled peptides revealed that the most common stapling positions are
i,
i + 4 or
i, i + 7, accounting for 82.4% and 16.5%, respectively (
Figure 2B). Another observation derived from this analysis is about the most popular stapling strategy, which is all-hydrocarbon stapling, accounting for 89.5% (
Figure 2C)
[37].
Figure 2. Classification of AMPs deposited on the Data Repository of Antimicrobial Peptides (DRAMP) describing (A) type of AMP, (B) most common positions for stapled AMPs and (C) stapling strategy applied.
2.1. All-Hydrocarbon Stapling
All-hydrocarbon stapling was first established by a research group led by Prof. Gregory Verdine
[22]. Inspired by the development of the chemistry for olefinic cross-linking of helices, which is performed through
O-allyl serine residues via ruthenium-catalyzed ring-closing metathesis (RCM), Verdine’s group designed non-natural amino acids having either
R or
S stereochemistry at the α-carbon loaded with tethers of different lengths, followed by RCM and forming the desired clamp at determined peptide positions (
Figure 3)
[22][23][38]. The olefin tether bridge used for helix stabilization was considered similar to an actual staple, which explains why the term “staple” was introduced in this technique. Stapled helix stabilization relies on the tether length and absolute configuration at the α-carbon position. For this reason, these parameters have been intensely studied. Additionally, it is possible to constrain helices with more than one stapling, independent staples (
Figure 1D) or via spiro-bicyclic ring connection, which is also called stitching (
Figure 1E)
[9][39][40][41].
Figure 3. Cross-linking mediated by ruthenium-catalyzed ring-closing metathesis (RCM) between non-natural amino acids presenting S (A) and/or R (B) stereochemistry forming the staple at the desirable position within the α-helix structure.
Regardless of having both single and double stapling strategies available, single stapled peptides stand out in academic research
[26][36][42][43]. The success of both strategies has shown to be peptide-dependent, as an increase in the helical content and improved antimicrobial activity is a premature correlation and not always observed to be true. Additionally, peptide stapling is often related to increased hemolytic activity, thus highlighting the importance of fine-tuning stapled AMP candidates.
The natural pro-inflammatory chemokine involved in response to microbial infections, CXCL10, is known to be active against different pathogens. However, there is evidence suggesting that CXCL10 have distinct antimicrobial domains. Taking this into account, Crawford et al.
[38] mapped the chemokine structure and designed a series of nine overlapping peptides to test against a range of Gram-negative and positive bacteria. Among the derivative peptides, peptide P9, which was originated by the C-terminal α-helix of CXCL10, demonstrated by CD spectroscopy that it requires a specific conducive environment to adopt the native α-helix conformation. That way, in order to promote the secondary structure and enhance the associated antimicrobial activity, P9 was submitted to a stapling at positions 8 and 12, being substituted for the non-natural amino acid (S)-pentenyl alanine. As a result, a great improvement in bactericidal activity was observed, despite the resulting increase in cytotoxicity
[44].
All-hydrocarbon stapled analogues of the naturally occurring peptide polybia-MPI (MPI) from the venom of the
Polybia paulista wasp
[45] were synthesized and had their chemical and biological properties evaluated by Luong et al.
[43]. The three analogues (MPIS, MPIS-D8N, and MPIS-Q12K) have an oct-4-enyl hydrocarbon staple at the
i,
i + 4 positions, tethering amino acid residues 6 and 10, whereas MPIS presented only the stapling.
As a chemical modification, MPIS-D8N and MPIS-Q12K also had single amino acid substitutions. In MPIS-D8N, aspartic acid (D) in position 8 is substituted by asparagine (N), whereas glutamine (Q) 12 is substituted by lysine (K) in MPIS-Q12K. As a result, there was an increase of at least 3-fold in the α-helical contents of all analogues compared to their parent peptide, MPI. However, the antimicrobial activity was not necessarily better enhanced. All three analogues showed improved antibacterial activity between 7- and 23-fold against Gram-positive bacteria; still, there was no improvement for the inhibition of Gram-negative bacteria. The hemolytic assay results revealed that all stapled peptides displayed an increase in hemolytic activity as follows: 12-fold for MPIS-D8N and approximately 4-fold for MPIS and MPIS-Q12K. In the last experiment, MPIS and MPIS-Q12K were submitted to a trypsin digestion assay and showed a 68-fold cleavage decrease
[43].
As described above, hemolysis is an obstacle that must be overcome. For that reason, researchers seeking to develop and design stapled peptides should put effort into finding a well-balanced set of parameters, not only focusing on increasing hydrophobicity and helical content. In this line, Stone et al.
[36] used the previously studied synthetic antimicrobial peptide 6K-F17 (KKKKKK-AAFAAWAAFAA-NH
2)
[46][47] to generate four stapled analogues; all of them clamped at the same position, but three of them did so with amino acid substitutions. The stapled S-6K-F17 with no amino acid substitution showed a higher α-helix formation tendency, increased antimicrobial activity against
E. coli BL21, and a more rapid killing mechanism compared to the non-stapled version (6K-F17) and the other three stapled analogues with amino acid substitutions (S-6K-F17-2G, S-6K-F17-3G and S-6K-F17-3GN). Nevertheless, the observed hemolytic activity of S-6K-F17 was much higher, with an increase varying between 4- and 154-fold. However, the three stapled analogues with higher-polarity amino acid substitutions, namely glycine and/or asparagine, despite not presenting so much helicity, were less toxic to mammalian cells and had only minor antimicrobial activity shifts, 0- to 3-fold. That observed outcome is probably due to a better hydrophobicity balance and selectivity for prokaryotic membranes, since adding staples increases non-specific membrane interactions. These results highlight those constraining helixes, and although they have a high tendency to improve antimicrobial activity, they should not be the only focus in staple studies. Adding polar amino acids, which are unfavorable for helix formation and decrease the helicity index, helps with selectivity and thus with toxicity aspects.
Proteolytic stability is a crucial characteristic for AMP development. Although stapled peptides usually present higher resistance to enzymatic degradation, this property is also dependent on the composition of peptide amino acid residues. Aiming to understand the influence of stability, helicity and antimicrobial activity against pathogenic fungi using an all-hydrocarbon strategy, Zheng et al.
[42] developed single-stapled analogues at
i,
i + 4 positions using aurein 1.2, a peptide secreted by the Australian bell frog
Litoria aurea [48]. After CD spectroscopy analyses, most stapled peptides showed an increase in their helical content. However, this increase did not necessarily lead to an improvement in antifungal activity against
Candida spp. strains. Four stapled analogues showed better antifungal activity and were selected to undergo a protease stability test by chymotrypsin degradation. Data indicate better proteolytic protection within all selected analogues compared to aurein 1.2. However, among these analogues, one presenting a higher helicity index showed lower proteolytic stability. Therefore, the results indicate that, despite achieving a boosted aurein 1.2 analogue, improving a peptide’s antimicrobial potency and stability can be more complex than just constraining helices
[42].
Besides all the difficulties that could be found, stapling continues to be a promising strategy for AMP rational design. Hirano et al.
[26] decided to improve the potential of a peptide 1 drug, a previously reported Magainin 2 (Mag 2) derivative, by designing and synthetizing 8 side chain stapled peptides, numbered from 2 to 9, in order to evaluate their preferred secondary structures, and their antimicrobial and hemolytic activities. These peptides were synthesized by introducing a suitable (S)-pentenyl alanine and/or (R)-octenyl alanine at the
i,
i + 4 or
i, i + 7 positions, which would lead to side chain stapling. As a result, peptides 2 to 6, presenting
i,
i + 4 stapling, showed improved antimicrobial activity, while peptides 7 to 9, presenting
i, i + 7 stapling, had decreased or indifferent shifts in antimicrobial activity. The investigation of hemolytic activity showed that among all the stapled peptides, only peptide 2 demonstrated hemolytic activity at much higher concentrations compared to MIC, estimated as being 16× higher for gram-positive
S. aureus and 32× higher for gram-negative
E. coli and
P. aeruginosa. All the other peptides presented hemolytic activity at concentrations as low as MIC or even lower, and peptide 8 was the one with the lowest hemolytic concentration. In addition, electrophysiological measurements were performed, revealing that peptide 2 showed higher scores against the DOPE/DOPG membrane than against the DOPC membrane. These results indicate a pore-formation preference for bacterial mimetic than for mammalian mimetic membranes.
Understanding the possible drawbacks related to characterizing stapled peptides, the importance of studies that pursue more complex and rich evaluations is clear, especially those taking advantage of computational approaches. Thus, it can be possible to generate better knowledge and insights on how to improve stapled peptides’ therapeutics and competitiveness.
2.2. Alternative Stapling Strategies
Although it is the most used stapling technique, hydrocarbon stapling is only one of the many approaches available. The term “stapling” has become a general term to refer to cyclic peptides, especially those that aim to constrain helices
[19]. Alternatively, there are one-component approaches such as disulfide or amide bonding, and two-component methodologies, including thioether formation and alkylation of Lys residues
[17]. In this perspective, Liu et al.
[49], also using polybia-MPI as the original counterpart, generated the stapled analogues C-MPI-1 and C-MPI-2 by a triazole stapling technique. From these, only C-MPI-1 presented a slight α-helical structure. Even though C-MPI-1 showed higher helicity content, there was little or no enhancement in antimicrobial activity against Gram-positive and Gram-negative bacteria and worse hemolysis results compared to MPI.
Similarly, the cationic AMP OH-CM6, derived from the natural cathelicidin peptide OH-CATH30 from the king cobra
Ophiophagus hannah was used to generate stapled analogues by N-alkylation of lysine
[50]. In this strategy, the lysine residues on the hydrophilic face of the native peptide were clamped by an alkyl linker at
i,
i + 4 and
i, i + 7. During this process, ten stapled peptides were generated and screened for antimicrobial activity against Gram-positive and Gram-negative bacteria, including clinical isolated strains. The best candidate (peptide 10) was chosen after antimicrobial screening. From this selected peptide, different analogues were synthesized with an aralkyl linker. One of these peptides (peptide 12) presented antibacterial activity against methicillin-resistant
S. aureus in concentrations comparable to Vancomycin. In addition to that, despite the toxicity observed in the assay against the HEK 293T cell line, the concentration was 32-fold higher than the observed minimum inhibitory concentration, showing an interesting therapeutic index. Moreover, the α-helical content was around 45% and the proteolytic stability was better than other analogues tested, namely the linear and one of the stapled analogues, indicating a promising molecule for further investigation.