SEARCH

SEARCH BY CITATION

Keywords:

  • α-peptoids;
  • antibiotic resistance;
  • antimicrobial peptides;
  • β-peptide;
  • drug design;
  • peptidomimetics

Abstract

  1. Top of page
  2. Abstract
  3. Peptidomimetics
  4. β-Peptide Structure
  5. Antibacterial and Hemolytic Activity of β-Peptides
  6. Peptoid Structures
  7. Antibacterial Activity of Peptoids
  8. Therapeutic Potential of Peptide or Peptide Related Molecules
  9. Conclusion
  10. Acknowledgment
  11. References

The field of drug discovery and development has seen tremendous activity over the past decade to better tackle the increasing occurrence of drug-resistant bacterial infections and to alleviate some of the pressure we put on the last-resort drugs on the market. One of the new and promising drug candidates is derived from naturally occurring antimicrobial peptides. However, despite promising results in early-stage clinical trials, these molecules have faced some difficulties securing FDA approval, which can be linked to their poor metabolic stability. Hence, mimetics of these antimicrobial peptides have been suggested as new templates for antibacterial compound design, because these mimetics are resistant against degradation by proteases. This review will discuss the structural features of two different types of mimetics, β-peptides and α-peptoids, in relation to their antibacterial activity and conclude on their potential as new candidates for bacterial intervention.

Conventional antibiotics target bacteria by destroying their cell wall or by blocking biosynthetic pathways necessary for the bacteria to survive. Unfortunately, there has been a rapid increase in clinical bacterial strains that have acquired mutations, which render them resistant against one or several conventional antibiotics, seriously limiting the treatment options (1). Consequently, there is a need to search for new antibacterial agents with a new mode of action to overcome problems related to bacterial resistance. With that perspective in mind, and results from mapping of bacterial genomes, it should be possible to find new treatment strategies that target elements essential for bacterial survival. One such approach could be tailored on the characteristics of antimicrobial peptides (AMPs), which are produced by a wide range of organisms as part of their defense strategy against microbes. Today, more than 1000 AMPs have been isolated and described (2) from a wide range of animals: single-celled micro-organisms, invertebrates, plants, amphibians, birds, fish, mammals including humans [for review, see (3–5)]. The most common AMPs are cationic and amphipathic molecules consisting of <40 amino acid residues. The frequent mode of action of these peptides is in part related to cell membrane permeabilization and lysis. Despite differing in their primary and secondary structures, AMPs share some common structural patterns. They commonly fold up into different amphipathic structures, separating the hydrophobic and the cationic parts (6–8). This structure enables the peptides to adhere to the surface of the bacteria through electrostatic interactions, resulting in the rapid lysis of a broad range of bacteria, which decrease the likelihood of developing resistance (9). As a result, much attention has been given to AMPs; they are thought to be a novel class for antibacterial treatment, and rightfully so. There are several examples of naturally occurring AMPs that have been slightly modified before entering into human clinical testing (10–12). The success of these peptides as drug candidates has been questioned, and much of the criticism is related to their low bioavailability and lack of metabolic stability in vivo (13,14). This has led to the idea of using the AMPs as templates for synthetic mimics, modified in different ways to improve these properties. This review will focus on two types of antibacterial peptidomimetics, namely β-peptides and α-peptoids, discussing their structural requirements for antibacterial activity and looking into their promise as new antibacterial drug candidates.

Peptidomimetics

  1. Top of page
  2. Abstract
  3. Peptidomimetics
  4. β-Peptide Structure
  5. Antibacterial and Hemolytic Activity of β-Peptides
  6. Peptoid Structures
  7. Antibacterial Activity of Peptoids
  8. Therapeutic Potential of Peptide or Peptide Related Molecules
  9. Conclusion
  10. Acknowledgment
  11. References

Mimics of AMPs are collectively called peptidomimetics. The types of modifications that are introduced are generally modeled after the structural requirements which are known to influence AMP activity. Attempts to conserve features like positive charge and amphipathicity are made to ensure the antibacterial activity of the mimetic compounds. The mimics are often constructed with a different backbone (i.e., not based on α-amino acids) or may carry dislocated side chains to overcome the low bioavailability and lack of metabolic stability found for traditional AMPs (15). This mimicry is highly advanced because of intense research interest over decades [for review (16,17)]; however, only the β-peptide and the N-substituted glycine (peptoid) strategy will be covered within this review. The antibacterial β-peptides were originally designed as short oligomers of β-amino acids (Figure 1A) mimicking the AMP magainin (Figure 2A) (18,19). Although β-peptides are still viewed as ideal oligomers for folding into α-helical structures (20), it has also been demonstrated that β-peptides can adopt a variety of different helical conformations (21–25): C12/C10/C12 hydrogen-bonding turn confirmations (26,27) in addition to antiparallel hairpin and sheet-like structures if the appropriate β-amino acids are added in the sequence to stabilize these conformations (Figure 2B) (28–30). Another promising mimic of antibacterial peptides is called peptoids (Figure 1A), and they are comprised of N-substituted glycine (31). Hence, the reactive side chain in the natural amino acid has been moved from the α-carbon to the backbone nitrogen (32), a modification resulting in higher metabolic stability (33) and low toxicity to mammalian cells (15). Similar to β-peptide research, one strategy in α-peptoid design has also been applied to AMPs with only a few point substitutions (34), while another strategy is based on mimicking the properties and the overall structural characteristics of an AMP (35). These strategies have demonstrated that properties like hydrophobicity, positive charge, and amphipathicity need to be imitated to ensure the antimicrobial activity of both β-peptides and α-peptoids (9,36). Naturally, these are also chemical properties that have proven to be crucial for traditional AMP activity, where the hydrophobic portion enables interaction with the phospholipid membrane of the bacteria, because cationic segments are attracted to the negatively charged bacterial surface (37–41).

image

Figure 1.  Subunits for design of peptidomimetic substances. Subunits for construction of an α-peptide, different β-peptides, and a peptoid (A) in addition to the different possible positions for side chain substitution. α-peptides; three substitution positions and two configurations. β-peptides; 5 substitution positions and four configurations (B). An extra methylene group in the β-peptide backbone enables ‘free’ rotation around an additional bond in the backbone (C) (45). To make the backbone more rigid, different β-amino acids can be incorporated (D).

Download figure to PowerPoint

image

Figure 2.  Hydrogen bond stabilization of different helical β-peptide conformations. The two-dimensional structure of magainin modeled from PDB 2MAG (93) using Swiss-PdbViewer (94) illustrates the amphipathic organization of the peptide in a lipid environment (A). Different hydrogen bond arrangements found in β-peptides (B, C), a list of the backbone constituents with the type of helix they can form and the number of residues per turn in these helices (D) (22,26,51,66,95,96). Hydrogen bond pattern in an α/β-peptide rigidified by the incorporation of ACPC (β-amino acids) in every second position. The 11-membered ring is indicated in yellow, while the 14- and 15-membered rings are depicted in purple (E).

Download figure to PowerPoint

β-Peptide Structure

  1. Top of page
  2. Abstract
  3. Peptidomimetics
  4. β-Peptide Structure
  5. Antibacterial and Hemolytic Activity of β-Peptides
  6. Peptoid Structures
  7. Antibacterial Activity of Peptoids
  8. Therapeutic Potential of Peptide or Peptide Related Molecules
  9. Conclusion
  10. Acknowledgment
  11. References

Although β-amino acids (Figure 1A) do exist as part of some natural occurring peptides isolated from prokaryotes and marine organisms (42), the general perception is that they are rare. Because of their extra methylene group, these β-amino acids are not recognized by traditional proteases, rendering mimics composed of β-amino acids intrinsically stable against degradation (43). Studies have also shown that mimics composed of alternating α- and β-amino acids (α/β-peptides) residues are very resistant toward degradation by trypsin, pronase, and chymotrypsin (44). Introduction of the extra carbon atom in the backbone also creates room for an extra side chain. Thus, β-amino acids can be either mono- or double-substituted, and a mono-substituted β-amino acid can have the side chain positioned at two different carbon atoms (26). To distinguish between the side chain orientations, the R1 and R2 positions of the side chain give rise to β2-amino acids and the R3 and R4 positions are called β3-amino acid (Figure 1B). Furthermore, a peptide built exclusively from β3-amino acids is called a β3-peptide. Likewise, a peptide using β2-amino acids as the sole building block is called a β2-peptide (26). In addition to this, it is also crucial to remember that introduction of an extra methylene group in the backbone adds an extra free-rotating C-C bond, resulting in torsional degrees for freedom defined by ϕ, ψ, and θ (Figure 1C) (45). This was originally believed to obstruct the formation of stable secondary structures. However, later studies showed that peptides made from β-amino acids can indeed form a variety of stable secondary structures (26,46,47).

Although β-peptides can form a variety of structures, the main focus for design of AMP mimics has been on creating β-peptides that can adopt different helical conformations (Figure 2B–D). This strategy has been questioned by some authors who have argued that the helical conformation is less important in inducing antibacterial activity but rather plays a greater role in hemolytic activity (48,49). Despite this discrepancy, it is still generally accepted that an amphipathic structure is a key factor in inducing antibacterial activity. In light of this, many studies performed with β-amino acids were attempting to mimic the chemical features of magainin (Figure 2A), and the ability of these β-peptides to form helical structures has been tested numerous times. Results from circular dichroism spectroscopy studies have indicated that β-peptides of 6 to 12 residues are able to form helical structures in an organic environment (20,22,26), whereas comparable peptides consisting of α-amino acids require 15 residues to form a stable helix (26). This demonstrates that the extended backbone element in the β-peptides rather increases, than prevents the tendency of these peptides to form helical structures, a phenomenon that is also influenced by the positioning and the chemical nature of the side chain(s) (50). An amphipathic β-peptide conformation can also be achieved using different primary templates, as these β-peptide helices can be formed with varying numbers of residues per turn (Figure 2D).

Although β-peptides are capable of forming helical configurations when introduced to lipids or organic solvents, it is also important to note that these peptides are unable to form a helical structure in aqueous solution (20,51,52). This observation demonstrates in many ways how closely related β-peptides are to traditional AMPs (α-peptides), which also tend to undergo a conformational change upon interaction with a lipid matrix (53–55). It has also been suggested that because peptides built from β-amino acids display resistance toward degradation; they could offer a new design strategy for peptidomimetics of traditional protease-sensitive antibacterial peptides. Hence, synthesis of several different β-peptides has been conducted in search for a structure–activity relationship between peptide helicity, high antibacterial, and low hemolytic activity.

Antibacterial and Hemolytic Activity of β-Peptides

  1. Top of page
  2. Abstract
  3. Peptidomimetics
  4. β-Peptide Structure
  5. Antibacterial and Hemolytic Activity of β-Peptides
  6. Peptoid Structures
  7. Antibacterial Activity of Peptoids
  8. Therapeutic Potential of Peptide or Peptide Related Molecules
  9. Conclusion
  10. Acknowledgment
  11. References

One study performed by Gellman’s group has demonstrated promising results with a 17-residue magainin analog consisting of hydrophobic ACPC and cationic APC subunits (Figure 1D). This 17-meric β-peptide (ACPC2-[APC-ACPC-APC-ACPC2]3) adopted a 12-helix with approximately 2.5 residues per turn (24,56), which made it possible to design an amphipathic oligomer with the cationic APC residues on one side and the hydrophobic ACPC residues on the opposite side (18,57). The minimal inhibitory concentration (MIC) of the most potent 17-meric β-peptide (ACPC2-[APC-ACPC-APC-ACPC2]3) was measured to 3.2 and 6.3 μg/mL against S. aureus and E. coli (18), respectively, making it slightly more potent compared to magainin (19). The study clearly demonstrated the importance of an amphiphilic helix for the antimicrobial activity of β-peptides. Furthermore, certain ratios of charged to hydrophobic residues were also demonstrated to give more potent perturbation of the bacterial membrane (57), suggesting a similar mode of action for the 17-meric β-peptides as for many traditional cationic AMPs (58–61). Although the hemolytic activity of two of these 17-meric β-peptides (ACPC2-[APC-ACPC-APC-ACPC2]3 and ACPC2-[AP-ACPC-AP-ACPC2]3) was not conclusively low (18,57), these peptides should be viewed as good candidates for further research, as they exhibited a significant therapeutic index. The hemolytic activity for the reviewed peptidomimetics is mainly assessed after only 1 h of incubation. Although it is fair to argue that destabilization, perturbation, or lysis of lipid membranes appears to be a rapid process if viewed as direct antibacterial killing (4,62), we believe that a hemolytic activity measured over a longer period of time (e.g. 24 h) would have given a stronger (and perhaps more reproducible) toxic effect. It may also be of importance to monitor hemolytic activity for an extended period of time as these peptidomimetics are designed to be resistant to proteolytic degradation and would thus linger around throughout the whole experiment (33,63).

Similar studies have been conducted on magainin analogs constructed of repeated elements of H-(β3Ala-β3Lys-β3Val)n-NH2 with either 12 or 15 residues (n = 4 or 5) adopting amphipathic 14-helices (52). Both mimics elicited antibacterial activity with MIC values ranging from 12 to 19 μm (20–26 μg/mL) against E. coli, which is slightly weaker than the 17-meric β-peptide (ACPC2-[APC-ACPC-APC-ACPC2]3) reported by Porter et al. (18). However, the H-(β3Ala-β3Lys-β3Val)4-NH2 peptide demonstrated a significant higher selectivity for bacterial membranes than the 17-meric β-peptide (ACPC2-[APC-ACPC-APC-ACPC2]3), with a therapeutic index (calculated from the hemolytic activity, HD50 values) of 48 and 10 against E. coli, respectively (18,52). This improved selectivity can be traced back to substitutions made in the peptide sequence. H-(β3Ala-β3Lys-β3Val)4-NH2 is derived from H-(β3Val-β3Lys-β3Leu)4-NH2, which is far more hydrophobic and has virtually no selectivity for bacterial membranes over red blood cells (20), indicating that a lower hydrophobicity may lead to reduced hemolytic activity without destroying the antibacterial activity (20,52).

The same study of H-(β3Val-β3Lys-β3Leu)4-NH2 demonstrated that this peptide adopts a 14-helical conformation (Figure 2C,D) (20) and has later inspired design of two series of cationic 9- and 10-residue amphipathic β-peptides for evaluation of the correlation between 14-helical stability, antibacterial, and hemolytic activity (48). To induce 14-helical stability, the rigid ACHC was incorporated (Figures 1D and 2E). The results show no correlation between the propensity of the peptide to form a 14-helix in aqueous solutions and their antibacterial activity. However, a greater tendency for 14-helices gave rise to a slightly higher hemolytic activity, which could suggest that helicity is more essential for hemolysis than for antibacterial activity (48). Most of the mimetics from this study elicited MIC values in the range of 0.6–10 μm (0.8–12.5 μg/mL), but as mentioned earlier, they did also possess high hemolytic activity (48). Furthermore, enantiomers of the tested mimics also demonstrated a similar activity profile as their homologs, which leads to the suggestion that the mode of action of these β-peptides does not rely on interactions with chiral receptors (48). Similar results have also been shown in studies on several traditional AMPs (64).

Other studies of amphipathic, 14-helical β-peptides have demonstrated results which are less encouraging. Hamuro et al. failed to achieve a reasonable correlation between antibacterial activity and hemolysis. From a library of 6- to 18-residue mimetics corresponding to repeated sequence of β3Val/Leu-β3Lys-β3Leu, only one 12-residue mimetic H-(β3Val/Leu-β3Lys-β3Leu)4-NH2 elicited antibacterial activity, but this compound was also very hemolytic (20). Similar results have been reported by Arvidsson et al. (65) who used the repeated sequence β3Ala-β3Lys-β3Phe to design 9- to 18-residue β-peptides in the 14-helical conformation. Their H-(β3Ala-β3Lys-β3Phe)4-NH2 peptide demonstrated detectable inhibition of bacterial growth; however, this was limited to the Gram-negative E. coli and K. pneumonia at the MIC range of 32–64 μg/mL (65). However, this peptide also showed the highest hemolytic activity (15% at 300 μm). Despite this, the most surprising observation from this study was clearly the fact that only two Gram-negative bacteria were inhibited by the mimetic, while Gram-positive strains like Enterococcus faecalis, S. aureus, Streptococcus pneumoniae, and Pseudomonas aeruginosa were unaffected. These studies could indicate that both the H-(β3Val/Leu-β3Lys-β3Leu)4-NH2 and the H-(β3Ala-β3Lys-β3Phe)4-NH2 sequence are too hydrophobic because increased hydrophobicity has been suggested to cause increased hemolysis (20,65). To address this, a nine-residue β23-peptide [H-β3Phe-β2Val-β3Phe-(β2Leu-β3Lys)22Leu-β3Phe-NH2] folding into a C12/C10-helix conformation was made with the intention of reducing the hydrophobic sector in the helix (Figure 2B,D) (66). The new peptide showed both an increased spectrum and increased antibacterial activity, improving the therapeutic index although this optimized β-peptide was less active than many of the more potent candidates discussed earlier. From these limited data, it is also unclear whether the improved antimicrobial activity of this C12/C10-helical sequence is a result of the altered three-dimensional hydrophobicity profile of the peptide or a result of the change in the peptide primary sequence.

A different strategy in creating peptidomimetics is the mixing of α- and β-amino acids. The most successful example of this is a series of dimers constructed of different lipophilic β2,2-amino acids that are combined with a C-terminal αArg. Structural optimization through employment of the pharmacophore model for short AMPs (67) has resulted in candidate peptidomimetics demonstrating MIC activities against both methicillin-resistant and -sensitive S. aureus in a range from 2 to 3 μm while retaining a meaningful therapeutic index (68). However, interpretation of this should be carried out with some caution, as some may argue that these peptidomimetics should be considered as small molecules rather than peptide derivatives. Regardless, their activity is quite strong and they could easily be viewed as interesting candidate molecules for future clinical optimization.

In a different α/β-peptide study, three compounds were constructed from αLeu, αLys, and the two β-amino acids ACPC and APC (Figure 1D) (44). Their primary structure was designed such that one of the peptides would be globally amphipathic in an 11-helix conformation (three residues per turn) and one would be globally amphipathic in a 14/15-helix (4.5 residues pr turn) conformation (Figure 2E). The last α/β-peptide was not globally amphipathic in either helix conformation, thereby having the lipophilic and cationic residues dispersed around the axis of the helix. Surprisingly, the α/β-peptide which was globally non-amphipathic [ACPC-αLeu-APC-(αLys-APC-αLeu-ACPC)2-(αLeu-ACPC)2] demonstrated the strongest antibacterial activity with MIC values of 6.3 and 12.5 μg/mL against E. coli and S. aureus, respectively (44). Further analysis demonstrated that this in fact was a result of the higher density of charged residues in the first portion of the peptide (longitudinal amphipathicity) and that by dispersing the charged residues more evenly along the entire helix, the antibacterial activity was lost (49). Additionally, this peptide did not elicit any hemolytic activity up to a concentration of 50 μg/mL (44,49), although the significance of this could be questioned because there is virtually no gap between the MIC values and the hemolytic concentration tested. However, a suggested antibacterial mode of action of these non-amphipathic peptides could be that they adopt irregular non-helical amphipathic conformations upon interaction with the lipid membrane. This would require great structural flexibility which could explain why mimics using ACPC have been demonstrated to be antibacterial, while mimics with more rigid ACHC (Figure 1D) have failed to inhibit bacterial growth (48,49). The overall results clearly indicate that there are elements at a primary structure level that influence the overall antimicrobial outcome of these sequences, an observation supported by years of research on traditional AMPs.

Peptoid Structures

  1. Top of page
  2. Abstract
  3. Peptidomimetics
  4. β-Peptide Structure
  5. Antibacterial and Hemolytic Activity of β-Peptides
  6. Peptoid Structures
  7. Antibacterial Activity of Peptoids
  8. Therapeutic Potential of Peptide or Peptide Related Molecules
  9. Conclusion
  10. Acknowledgment
  11. References

N-substituted glycines can be viewed as amino acids, where the side chain is attached to the amine nitrogen instead of the α-carbon, and oligomers of these building blocks are called α-peptoids (Figures 1 and 3). Similar N-substituted β-glycines will give rise to β-peptoids. The conformational change in the N-substituted glycines makes the α-carbon achiral so that both peptoids are less restricted in their spatial conformation. Neither peptoids can form intramolecular hydrogen bonds through backbone–backbone interactions, because of the lack of amide protons, that help peptides stabilize both α-helical structures and β-sheet conformations (69–71). However, the same backbone structure renders the peptoids highly resistant to proteases (15,71).

image

Figure 3.  Common peptoid monomers. Incorporation of many bulky residues and l-proline (Npro) will help stabilize an α-helical structure. Nspe, N-(S)-(1-phenylethyl)glycine; Nsch, N-(S)-(1-cyclohexylethyl)glycine; Nphm, N-(phenylmethyl)glycine; Nmeb, N-(4-methylbenzyl)glycine; Nbut, N-(butyl)glycine; Nsdp, N-(S)-(2,3-dimethylbutyl)glycine; Nssb, N-(S)-(sec-butyl)glycine; Nsmb, N-(S)-(1-methylbutyl)glycine; NLys, N-(4-aminobutyl)glycine; NPro, l-Proline; Nmna, N-(1-naphthalenemethyl)glycine; Nsna, N-(S)-(1-naphthylethyl)glycine.

Download figure to PowerPoint

Extensive studies on traditional AMPs have concluded that their ability to form an amphipathic structure upon interaction with lipid membranes is one of the driving forces behind their antimicrobial potential (4). Consequently, much research has focused on designing α-peptoids mimicking an α-helical structure (72,73). Because peptoids lack the ability to form hydrogen bonds that may stabilize the helical structure through backbone–backbone interactions, considerations on the peptoid primary structure are important during the design process. Incorporation of about 50% bulky α-chiral monomers or aromatic side chains has proven necessary to obtain a stable α-helical secondary α-peptoid structure (Figure 3) (73). The incorporation must be periodic so that the aromatic side chains can be evenly distributed along a sector in the α-helical structure (73). To obtain this periodicity, α-peptoids are often designed with a core sequence (NAaa-NBbb-NCcc)n repeated n times, where N designates the amide nitrogen and Aaa, Bbb, and Ccc designates the three-letter code for the attached amino acid side chain (69). α-Peptoid helicity is also strengthened by insertion of an α-chiral residue at the C-terminal end, and the longer the peptoid oligomer is, the more stable the helical structure will be (73). Therefore, when aiming to design a bioactive peptoid with helical secondary structure, the sequence should consist of cationic residues (e.g., NLys), hydrophobic aromatic residues (e.g., NPhe), and hydrophobic aliphatic residues (e.g., NIle) (69). Incorporation of residues with these properties gives rise to a helix with the desired amphipathic and hydrophobic features, which are important for the α-peptoid to act on bacteria, as mentioned earlier.

Although most emphasis in research has been laid on the secondary structure of α-peptoids, studies have also demonstrated that a 15-mer α-peptoid may be capable of forming a tetramer complex, thereby mimicking the tertiary structure of proteins (74). This discovery increases the field of possibilities for using peptoid mimics in designing other biologically relevant oligomers. While different non-covalent interactions are proposed to play a role in the folding process (e.g., steric repulsion) (75,76), α-peptoid folding is still not completely understood despite intense research (70).

Antibacterial Activity of Peptoids

  1. Top of page
  2. Abstract
  3. Peptidomimetics
  4. β-Peptide Structure
  5. Antibacterial and Hemolytic Activity of β-Peptides
  6. Peptoid Structures
  7. Antibacterial Activity of Peptoids
  8. Therapeutic Potential of Peptide or Peptide Related Molecules
  9. Conclusion
  10. Acknowledgment
  11. References

There are several strategies for mimicking an AMP using α-peptoid subunits. The most intuitive strategy would be to use exactly the same sequence of the side chains in the α-peptoid as in the AMP. However, this rarely gives the same activity as the AMP because the hydrogen-bonding properties of the backbone nitrogen are absent in the peptoid (70). Another approach would be to substitute single amino acid residues with α/β-peptoid residues in an AMP to retain as many of the original peptide properties as possible. This will potentially increase the protease stability of the compound, which is the overall goal of many mimetic designs. However, even modest modifications may have drastic effects on the overall structure of the new compound. Substitutions with α-peptoid subunits have been performed in melittin to assess whether this could be used in design of melittin analogs with diminished or completely removed hemolytic and cytotoxic activity while retaining the antibacterial properties. When substituting three leucine zipper residues in melittin with α-peptoid residues NLeu, NPhe, and/or NLys, the antibacterial activity was retained and three out of four α-peptoids demonstrated MIC values between 1 and 4 μm against Gram-negative E. coli and P. aeruginosa and against Gram-positive B. subtilis, S. epidermis, and S. aureus. However, these substitutions prevented the sequence from folding up into an α-helical structure, thus diminishing the cytotoxic and hemolytic activity (>100 μm) (34). The results are in good correlation with studies of traditional peptide analogs of melittin, which have demonstrated that substitutions in the leucine zipper can disturb the α-helical conformation of the peptide and that this region is essential for the hemolytic activity of the peptide (77). The α-peptoid-substituted melittin analogs further demonstrated antimicrobial activity against four antibiotic-resistant clinical isolates with MIC values between 2 and 8 μm (34).

Another study investigating the effect of these mixed peptidomimetics used a library of 5- to 10-mers made of α-amino acids (lysine) in combination with α-peptoid subunits. In general, the Lys-peptoid hybrids were more active against Gram-positive S. aureus than against Gram-negative E. coli (78). To verify the clinical relevance of these findings, efficacy against a selection of clinical bacterial strains was also assayed, demonstrating a significant effect of several compounds against all (i.e., methicillin-resistant S. aureus, vancomycin-resistant E. faecium, P. aeruginosa, S. typhimurium) but one species (K. pneumoniae), which in general was hard to inhibit (9). The best Lys-peptoid hybrid was a 9-mer compound Nmna-Nmeb-Nmna-Nbut-(αLys)5 (Figure 3) consisting of four α-peptoid residues and five Lys residues, which demonstrated a MIC of 3.1 μm or less against all the Gram-positive bacterial strains, and showed low hemolytic activity (9). The Lys-peptoid hybrids that contained the highest number of Lys residues showed the highest activity against both Gram-negative and Gram-positive bacteria, stating the importance of Lys residues in peptidomimetic design to reach the desired physicochemical properties (9,71). This is not particularly surprising, as numerous experiments on traditional antibacterial peptides have demonstrated that cationicity and bacterial killing to a certain extent go hand in hand. However, an interesting extension of this study by Ryge et al. would be to trade lysine for arginine to determine whether this could increase the antibacterial activity even further, as the guanidinium group of arginine often has been proven to be preferred over lysine (9,67).

An issue with incorporation of α/β-peptoid subunits dispersed throughout a native peptide sequence is the overall effect this will have on the structure of the compound. It is often perceived as suitable to create peptoids with an amphipathic helical structure. A strategy to ensure or increase the chance of this would be to design pure peptoids with alternating patterns of bulky and charged residues. This strategy was used by Patch and Barron (35) in a study to evaluate the structural requirements for antibacterial α-peptoid activity. In this study, a library of 12- to 17-mer α-peptoids was synthesized using the solid-phase sub-monomer method (32) to mimic magainin-2. A lysine-like α-peptoid residue was introduced in every third position separated by bulky α-chiral side chains (Figure 3) to give the α-peptoids the desired amphipathic and helical character. The activity of the α-peptoids was tested against Gram-negative E. coli and Gram-positive B. subtilis. Two of the peptoids [H-(NLys-Nspe-Nspe)4-NH2] and [H-(NLys-Nssb-Nspe)4-NH2] demonstrated MIC values of 0.82–9.9 μm (1.5–20 μg/mL), with most potent activity against B. subtilis, and very low hemolytic activity (35). Further research with these two α-peptoids has re-confirmed their bacterial selectivity over mammalian cells. More interestingly, it also demonstrated that the helical structure is of less importance for the antibacterial activity than previously anticipated (72). The authors concluded that the antibacterial activity of the α-peptoids is correlated with an overall cationic charge and moderate hydrophobicity (72). In light of their high therapeutic index, these α-peptoids are without doubt very promising lead candidates for further drug optimization.

Another very promising study was reported by Ng et al. (79) in the late nineties. Synthetic combinatorial α-peptoid libraries were tested for antibacterial activity, and the so-far shortest antibacterial α-peptoids were identified. Several α-peptoids of only 2 and 3 residues were reported to exhibit antimicrobial activity against Gram-positive and Gram-negative bacteria (79). The most promising candidate α-peptoid from this library demonstrated modest hemolytic activity and antimicrobial MIC values between 5 and 40 μm; however, it showed no particular antibacterial activity in vivo (80). Similarly, a positional scanning library containing more than 10 000 3-mer peptoids were tested for antimicrobial activity, but only moderate activities were found (81). No further research and improvements in α-peptoids from either of these libraries have been reported, probably as other longer α-peptoids have been proven to be far more potent. Moreover, it was later shown that for an α-peptoid to be significantly antibacterial, it needs to be at least 12 residues long (35).

The antibacterial efficacy of peptoids is generally assessed using traditional broth dilution methods (82) to evaluate its effect on planktonic growth of different bacterial species. However, much of the clinical challenges related to drug resistance and persistent bacterial infections are linked to bacterial biofilms (83). There are two general methods for antimicrobial intervention of biofilms, either destroying the biofilm or protecting the surface to prevent it from forming. The latter strategy has achieved much attention lately, and there are numerous examples of novel approaches; for example, tethering conventional drugs or AMPs to different surfaces (84–86) to create sterile surfaces on catheters, artificial joints, etc. that will inhibit biofilm formation. The results are certainly very promising, although it is apparent that the mode of action of the antibacterial peptide is somewhat changed once the peptides are linked to a surface (87). Similar studies have also been carried out with α-peptoid sequences that have been demonstrated to be potent antibacterial compounds against planktonic bacteria in earlier studies (35,72). By immobilizing the most antibacterial α-peptoids from these studies onto different model surfaces, it has also been demonstrated that it is possible for the given surface to compromise the membranes of the bacteria that adhered to it. However, bacterial adhesion to the surface was significantly higher once the surface was covered with immobilized α-peptoids compared to surfaces with no tethered peptoid (88). Several of these above-mentioned α-peptoids should be viewed as suitable candidates for further research on antibacterial agents.

Therapeutic Potential of Peptide or Peptide Related Molecules

  1. Top of page
  2. Abstract
  3. Peptidomimetics
  4. β-Peptide Structure
  5. Antibacterial and Hemolytic Activity of β-Peptides
  6. Peptoid Structures
  7. Antibacterial Activity of Peptoids
  8. Therapeutic Potential of Peptide or Peptide Related Molecules
  9. Conclusion
  10. Acknowledgment
  11. References

Polymyxin B/E and daptomycin are peptide-based antimicrobial drugs that are used for the treatment of Gram-negative and Gram-positive infections, respectively. In addition to this, there has been tremendous effort in moving traditional cationic AMPs or derivatives thereof into the clinic. As a reflection of this work, there are currently 12–15 peptides at different stages of clinical development (11,12,89). One of the more successful and advanced candidate peptides is Omiganan®, which has finished several phase III trials. However, despite successful prevention of Staphylococcus-related infections (90), Omiganan® has not been able to demonstrate superior protection compared to traditional anti-infectives. Other peptides have also struggled in clinical trails, most likely due to the fact that peptides generally have poor bioavailability and short half-life. These are characteristics that can be resolved by moving from peptides into peptidomimetics, and consequently, the peptidomimetic compounds in clinical trials have moved relatively rapidly up through clinical testing. Currently, there are three mimetics being investigated for antimicrobial activity, two of which have been inspired and derived from AMPs. In September 2010, PolyMedix (Radnor, Philadelphia, PA, USA) entered their defensin mimetic (PMX30036) into a randomized, blinded, active controlled phase II clinical trial, for evaluation of safety and efficacy in treating acute skin infections caused by S. aureus. Similarly, Lytix Biopharma AS (Tromsø, Norway) also has approval to commence phase II clinical testing of their Lytixar™ (LTX-109) for treatment of skin infections caused by Gram-positive bacteria. The last and less AMP-related mimetic compounds are Ceragenins, which has a sterol backbone with attached amino acids or other chemical groups (91). The most advanced ceragenins have demonstrated broad spectrum antibacterial activity against multi-drug resistant strains (92). Formulation of this has led to CeraShieldTM, which has demonstrated positive results in preclinical studies led by Cheragenix Pharmaceuticals (Denver, ON, USA). However, its future is dim because Ceragenix Pharmaceuticals filed for bankruptcy earlier this year.

For years, there have been rumors and doubts tied to traditional AMP research and advancement of this in clinical trials. Furthermore, many have spoken about how a single field only can take so much beating before it collapses. While highly opinionated, this sentiment also carries an aspect of financial reality. Regardless of the outcome for traditional AMPs, we firmly believe that antibacterial peptidomimetic compounds are here to stay. Additionally, peptidomimetic research may re-introduce optimism in the anti-infective peptide field. We are convinced that in the near future, we will see several new peptidomimetic compounds entering into preclinical testing, and are enthusiastically awaiting for their final evaluation.

Conclusion

  1. Top of page
  2. Abstract
  3. Peptidomimetics
  4. β-Peptide Structure
  5. Antibacterial and Hemolytic Activity of β-Peptides
  6. Peptoid Structures
  7. Antibacterial Activity of Peptoids
  8. Therapeutic Potential of Peptide or Peptide Related Molecules
  9. Conclusion
  10. Acknowledgment
  11. References

Several studies of pure and mixed peptidomimetics based on β-peptides and α-peptoids have demonstrated that structural flexibility and the ability to form an amphipathic structure is of importance for the overall antibacterial activity. The amphipathic characteristics are clearly more important than the specific nature of the secondary structure. It appears that the same physicochemical properties demonstrated to be crucial for antimicrobial activity of peptides also apply for peptidomimetic compounds. Hence, peptidomimetic design should be able to benefit from decades of AMP research. Several of the promising peptidomimetic compounds discussed herein would be ideal candidates for further in vivo testing and optimization studies.

References

  1. Top of page
  2. Abstract
  3. Peptidomimetics
  4. β-Peptide Structure
  5. Antibacterial and Hemolytic Activity of β-Peptides
  6. Peptoid Structures
  7. Antibacterial Activity of Peptoids
  8. Therapeutic Potential of Peptide or Peptide Related Molecules
  9. Conclusion
  10. Acknowledgment
  11. References