This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jenssen, H. Articles by Hancock, R. E. W. Search for Related Content PubMed PubMed Citation Articles by Jenssen, H. Articles by Hancock, R. E. W.

 Previous Article  |  Next Article 

Clinical Microbiology Reviews, July 2006, p. 491-511, Vol. 19, No. 3
0893-8512/06/$08.00+0     doi:10.1128/CMR.00056-05
Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Peptide Antimicrobial Agents

Centre for Microbial Diseases and Immunity Research, University of British Columbia, Lower Mall Research Station, 232-2259 Lower Mall, Vancouver, British Columbia V6T 1Z4, Canada

SUMMARY

INTRODUCTION

NATURAL DISTRIBUTION AND ACTIVITIES OF ANTIMICROBIAL HOST DEFENSE PEPTIDES

ANTIVIRAL ACTIVITY

Structural Requirements for Antiviral Peptides

Mode of Action of Antiviral Peptides

Blocking of viral entry by heparan sulfate interaction.

(i) Blocking of cell-to-cell spread.

Blocking of viral entry by interaction with specific cellular receptors.

Blocking of viral entry by interaction with viral glycoproteins.

Membrane or viral envelope interaction. (i) Viral envelope interaction.

(ii) Cellular membrane interaction.

Intracellular targets and host cell stimulation.

ANTIBACTERIAL ACTIVITY

Structural Requirements for Antibacterial Peptides

Mode of Action of Antibacterial Peptides

Membrane-permeabilizing peptides.

Peptides that do not act by membrane permeabilization.

ANTIFUNGAL ACTIVITY

Structural Requirements for Antifungal Peptides

Mode of Action of Antifungal Peptides

ANTIPARASITIC ACTIVITY

DEVELOPMENT OF ANTIMICROBIAL PEPTIDES FOR CLINICAL APPLICATIONS

CONCLUSION

ACKNOWLEDGMENTS

REFERENCES

Top Next References

INTRODUCTION

Top Previous Next References

 
A wide variety of organisms produce antimicrobial peptides as part of their first line of defense (90). Antimicrobial peptides are typically relatively short (12 to 100 amino acids), are positively charged (net charge of +2 to +9), are amphiphilic, and have been isolated from single-celled microorganisms, insects and other invertebrates, plants, amphibians, birds, fish, and mammals, including humans (159, 257). To date, hundreds of such peptides have been identified (88), indicating their importance in the innate immune system (89). The expression of these antimicrobial peptides can be constitutive or can be inducible by infectious and/or inflammatory stimuli, such as proinflammatory cytokines, bacteria, or bacterial molecules that induce innate immunity, e.g., lipopolysaccharides (LPS) (44, 86). Some of these peptides are potent antimicrobials. In contrast, the direct antimicrobial activity of others is largely evident in dilute media, and direct microbe killing is almost certainly prevented by physiological conditions, including high monovalent or moderate divalent cation concentrations, host proteases, polyvalent anions such as glycosaminoglycans (e.g., heparan sulfate), and low local peptide concentrations. Conversely these peptides are important effector molecules of the innate immune system (24, 30). They are able to enhance phagocytosis, stimulate prostaglandin release, neutralize the septic effects of LPS, promote recruitment and accumulation of various immune cells at inflammatory sites (58, 266), promote angiogenesis (129), and induce wound repair (36). Peptides of mammalian origin have also been demonstrated to have an active role in the transition to the adaptive immune response by being chemotactic for human monocytes (246) and T cells (40) and by exhibiting adjuvant and polarizing effects in influencing dendritic cell development (47). Although such peptides may have a direct effect on the microbe, such as by damaging or destabilizing the bacterial, viral, or fungal membrane or acting on other targets, they appear to be broadly involved in the orchestration of the innate immune and inflammatory responses (89). Thus, they are increasingly being referred to as host defense peptides. For example -defensins are almost certainly bactericidal at the high (mg/ml) concentrations found in neutrophil granules, but they probably act primarily as immunomodulators at the lower concentrations released by degranulation at inflammatory sites.

Despite their similar general physical properties, individual cationic peptides have very limited sequence homologies and a wide range of secondary structures with at least four major themes. The most prominent structures are amphiphilic peptides with two to four ß-strands, amphipathic -helices, loop structures, and extended structures (21, 87) (Fig. 1; Table 1). This review provides an overview of the (direct) antimicrobial functions of these peptides, with an emphasis on antiviral activity and an update on antibacterial, antifungal, and antiparasitic activities.

View larger version (27K): [in this window] [in a new window]   FIG. 1. Structural classes of antimicrobial peptides. (A) Mixed structure of human ß-defensin-2 (PDB code 1FQQ) (216); (B) looped thanatin (PDB code 8TFV) (156); (C) ß-sheeted polyphemusin (PDB code 1RKK) (202); (D) rabbit kidney defensin-1 (PDB code 1EWS) (165); (E) -helical magainin-2 (PDB code 2MAG) (76); (F) extended indolicidin (PDB code 1G89) (212). The disulfide bonds are indicated in yellow, and the illustrations have been prepared with use of the graphic program MolMol 2K.1 (132).

Top Previous Next References

Antimicrobial peptides produced by bacteria were among the first to be isolated and characterized (163). While they do not protect against infection in the classical sense, they contribute to survival of individual bacterial cells by killing other bacteria that might compete for nutrients in the same environment. Bacterial antimicrobial peptides, also called bacteriocins, are thought to be produced by many or most bacteria (128, 206) and are generally extremely potent compared with most of their eukaryotic counterparts. Their activities may be either narrow or broad spectrum, capable of targeting bacteria within the same species or from different genera. The bacteriocins constitute a structurally diverse group of peptides, and it was recently proposed that they be classified into two broad categories: lanthionine containing (lantibiotics) and non-lanthionine containing (43). Lantibiotics are characterized by the inclusion of the unusual amino acid lanthionine and the necessity for posttranslational processing to acquire their active forms. The most extensively studied lantibiotic is nisin, produced by Lactococcus lactis, which has been commonly used for nearly 50 years as a food preservative without significant development of resistance. It is also extremely potent, displaying activity against a variety of gram-positive bacteria at MICs in the low nanomolar range. These properties have prompted intense study of the mechanism of action of nisin, which is discussed later in this review. Other lantibiotics have also received attention due to their possible applications in the treatment of bacterial species which have developed antibiotic resistance. Mersacidin, a tetracyclic peptide that is produced by Bacillus spp. (38, 39), displays bactericidal activity against methicillin-resistant Staphylococcus aureus that is comparable to that of vancomycin, but without the development of cross-resistance (135).

In plants, it is widely believed that antimicrobial peptides play an important and fundamental role in defense against infection by bacteria and fungi. Observations to support this role include the presence and expression of genes encoding antimicrobial peptides in a wide variety of plant species investigated thus far, demonstrations of their bactericidal and fungicidal activity in vitro, and correlations between expression levels of peptides and susceptibility to a given pathogen or the extent of resistance of a particular bacterium to plant-derived peptides and its virulence. So far, only peptides with a ß-sheet globular structure have been identified in plants, with the two major and best-studied groups being thionins and defensins (reviewed in reference 72). Physiologically relevant concentrations of thionins are active against bacteria and fungi in vitro, and studies utilizing transgenic plants have shown that heterologous expression of thionins can confer protection against bacterial challenge (35, 59). Plant defensins display antibacterial and antifungal activities in vitro (245). Consistent with a defensive role, they are found in leaves, flowers, seeds, and tubers.

Since invertebrates lack the adaptive immune system found in vertebrate species, they are reliant solely upon their innate immune systems to counteract invading pathogens. Considering the extraordinary evolutionary success of this group of organisms, it is evident that invertebrate innate immune mechanisms are extremely effective. This has prompted intense studies of invertebrate species such as the arthropod fruit fly, Drosophila melanogaster, which has become a model system for the study of innate immunity and has led to the discovery of immune system strategies, such as pathogen recognition receptors (Toll-like receptors), that are conserved in higher organisms, including mammals. Numerous antimicrobial peptides have now been identified in invertebrates, and they are recognized as playing a key role in protection from pathogenic organisms. Indeed, the role of antimicrobial peptides and the regulation of their expression, including the signaling cascades involved, is well understood for Drosophila (111). Antimicrobial peptides are found in the hemolymph (plasma and hemocytes), in phagocytic cells, and in certain epithelial cells of invertebrates. They can be expressed constitutively, for example, in the hemocytes of marine arthropods such as shrimp, oyster, and horseshoe crab (11, 114), or induced in response to pathogen recognition, such as antifungal peptides in Drosophila (149). Among some of the prototypic invertebrate antimicrobial peptides are the -helical cecropins (fly hemolymph) and melittin (bee venom) as well as the ß-hairpin-like peptides tachyplesin and polyphemusin (horseshoe crab). The horseshoe crab-derived peptides possess some of the most potent antibacterial and antifungal activities observed, with MICs of <2 µg/ml (280). Interestingly, polyphemusin also displays activity against human immunodeficiency virus (HIV) (160). However, the most abundant group of antimicrobial peptides in invertebrates are the defensins, which are open-ended cyclic peptides with three or four disulfide bridges. The activities of invertebrate defensins can be divided according to whether their principal biological activity is directed toward bacteria or fungi (33).

Antimicrobial peptides have been isolated from a wide range of vertebrate species, including fish, amphibians, and mammals, indicating that, even in the presence of an adaptive immune response, these peptides have an important role in host defense. Direct microbicidal activity is associated with vertebrate antimicrobial peptides to various degrees under physiological conditions, and these activities likely contribute to the first line of defense, especially where they are found in very high concentrations, such as in the granules of phagocytic cells or the crypts of the small intestine (23, 25, 220, 265, 266). However, it is increasingly recognized that in addition to direct microbicidal activity, small cationic peptides perform critical immunomodulatory functions and may be involved in the control of inflammation, which serves to recruit a variety of other microbicidal mechanisms (24, 265, 266). Consistent with their role in direct and indirect antimicrobial defenses, antimicrobial peptides in vertebrates are found at sites that routinely encounter pathogens, such as mucosal surfaces and the skin, as well as within the granules of immune cells (24, 265, 266).

Amphibian skin glands have proven to be a rich source of antimicrobial peptides, with approximately 500 having been described to date as originating from this source. This represents a large proportion of the total number of reported antimicrobial peptides (207; http://www.bbcm.univ.trieste.it/tossi/pag1.html). The -helical magainins (272) are the prototypic amphibian antimicrobial peptides, with strong membrane-permeabilizing activity towards gram-positive and -negative bacteria, fungi, yeasts, and viruses. Structure-function relationships and the mechanism of action of magainin have been extensively studied, and these peptides have subsequently served as the template for development of the first (although ultimately unsuccessful) clinical antibacterial peptide treatment (74, 137). The broad antibacterial and antifungal activities of dermaseptins, isolated from the skin of South American frogs, have also been widely studied. In addition to their presence in the skin, amphibian antimicrobial peptides are produced in the mucosa of the stomach, indicating a role in protection from ingested pathogens. The best-characterized examples are the Asian toad peptides buforin and buforin II, which are generated by cleavage of the nucleosome protein histone 2A. A number of excellent reviews have covered this large group of antimicrobial peptides (33, 207, 224).

Cathelicidins are a large and diverse group of vertebrate antimicrobial peptides. They are characterized by a well conserved N-terminal segment (the cathelin domain) that is proteolytically cleaved to generate the mature, active peptide contained within the C terminus. Hence, most cathelicidins are stored in an inactive propeptide state, mostly within granules of circulating immune cells. Neutrophil secretory granules are the predominant source of cathelicidins, but they may also be expressed in mucosal surfaces in the mouth, lung, and genitourinary tract and in skin keratinocytes in inflammatory disorders, as is the case with human cathelicidin LL-37 (hCAP18) (66). Beyond the common N terminus, the structure of mature cathelicidins is diverse, with -helical, ß-hairpin, and proline/arginine-rich peptides all represented. The structural diversity within the cathelicidin family is also indicative of their apparently distinct functions, and they exhibit a diverse spectrum of microbicidal and immunomodulatory activities. Cathelicidins have been isolated from many mammalian species, such as mice, rabbits, sheep, horses, and humans. In some mammals, such as cattle, multiple cathelicidins are found in the body, indicating that they likely perform varied biological roles in host defense. One of the best characterized bovine antimicrobial peptides is BMAP-28, an -helical peptide which rapidly permeabilizes the membranes of a broad spectrum of bacteria and fungi at moderate concentrations in vitro (229), whereas the proline-rich bovine peptide Bac 5 shows selectivity for gram-negative bacteria under the same conditions (75). In contrast, only one cathelicidin is expressed in humans: LL-37 (hCAP18). Evidence in support of an early defensive role for LL-37 includes its upregulation in response to infection in the skin (54), as well as conditions in which a deficiency of LL-37 leads to chronic periodontal disease (204). In addition to direct microbicidal activity, LL-37 has important additional roles in host defense, including chemotactic properties and modulation of inflammatory responses (24, 271).

A second prominent group of mammalian antimicrobial peptides is the defensins (70, 221), cyclic peptides which are categorized into three subfamilies on the basis of the disulfide pairings between their six conserved cysteine residues (- and ß-defensins) or their macrocyclic nature (-defensins). As with cathelicidins, vertebrate defensins are synthesized as prepeptides which require proteolytic processing to their active peptide forms. The - and ß-defensins are widely distributed in vertebrate species, whereas -defensins have so far been identified only in Old World monkeys and apparently only in neutrophils and monocytes so far (244). Depending on the species, - and ß-defensins are found in the granules of neutrophils, macrophages, NK cells, intestinal Paneth cells and epithelial tissues, the skin, certain mucosal surfaces such as the respiratory passage and urinogenital tract, and many bodily fluids. Expression of the defensins may be constitutive, such as for human ß-defensin-1 (hBD-1) in most tissues, or inducible, such as for hBD-2, the expression of which in monocytes is upregulated following exposure to bacteria or LPS (56, 62). In vitro studies demonstrate that collectively, defensins possess generally weak microbicidal activities towards bacteria, fungi, and some viruses. While the bactericidal activities of most - and ß-defensins are antagonized by increasing salt concentrations (e.g., 100 mM monovalent and/or 2 mM divalent cations, which are found at many body sites), the high concentrations of -defensins that are found in some locations, particularly in the granules in phagocytic cells and in intestinal crypts, are thought to be sufficient to result in killing despite antagonism by salts (10, 71). -Defensins and hBD-3, on the other hand, retain their bactericidal activities in physiological salt conditions and also display antiviral activity in in vitro studies using HIV (42, 113). Transgenic and knockout experiments with mice have indicated a critical role for defensins in host defense. MMP-7-null mice, which lack all mature -defensins due to the loss of the protease required for proteolytic cleavage, display a reduced clearance of Escherichia coli and higher mortality rates upon challenge with Salmonella enterica serovar Typhimurium, indicating a important role in intestinal immunity (258). Conversely, knock-in of human HD-5 defensin into mouse Paneth cells conferred immunity to oral challenge using Salmonella enterica serovar Typhimurium (214). Importantly, in each case a correlation was observed between the antibacterial activity of the altered Paneth cell products in vitro and their protective abilities in vivo.

ANTIVIRAL ACTIVITY

Top Previous Next References

 
Representatives from all four structural classes of the cationic host defense peptides have shown the ability to inhibit viral infection. The spectra of viruses that are affected comprise primarily enveloped RNA and DNA viruses, with the exception of nonenveloped adenovirus (15, 102), feline calicivirus (164), and echovirus 6 (199). The antiviral activity of antimicrobial peptides often appears to be related to the viral adsorption and entry process (16) or is a result of a direct effect on the viral envelope (1, 208). However, it appears to be impossible to predict antiviral activity based primarily on secondary structures of peptides (Table 2). For example -helical peptides such as cecropins, clavanins, and the cathelicidin LL-37 have been shown to cause minimal or no herpes simplex virus (HSV) inactivation (18, 185, 269), while -helical magainins (Fig. 1), dermaseptin, and melittin have shown quite potent anti-HSV activity (Table 2) (1, 3, 16, 269). Conversely, ß-sheet peptides such as defensins, tachyplesin, and protegrins as well as the ß-turn peptide lactoferricin have all shown high activity towards HSV (Table 2) (4, 45, 120, 148, 225, 269, 270). It should be noted that within the different peptide subclasses, activity may vary considerably. For example, protegrin analogues lacking one or both disulfide bridges vary from highly active to inactive against HSV infections (269).

The creation of a series of lactoferricin analogues and study of their activity towards HSV revealed a relationship between the peptide net charge and its antiviral activity (120, 121). However, the spatial positioning of the charged amino acids seemed to be more important for antiviral activity than the actual net charge (120). For lactoferricin the nature of the aromatic amino acid appeared to be of minor importance for the antiviral activity, although its contribution to the secondary structure and thereby presentation of the charged residues might be crucial (121). Detailed studies on the influence of secondary structure domains on anti-HSV activity illustrated that the -helicity of a peptide could not explain its antiviral activity (122), thus implying that the presentation of the charged residues is of greatest importance with respect to anti-HSV activity (119). This is in accordance with results from Giansanti et al. (77) in a study on peptides derived from bovine lactoferrin and hen ovotransferrin. They concluded that the presence of hydrophobic and positively charged residues is critical but not sufficient for antiviral activity, and this may relate to different conformations adopted by these peptides in the context of the native protein (77).

-Defensins are rather rigid cyclic peptides from Old World monkeys. Analogues have been designed with a focus on Ile-to-Tyr or Arg-to-Tyr substitutions, in addition to the extensive use of D-amino acids, and have demonstrated the importance of the charge and spatial conformation of the peptides (270). Similarly, defined structures can provide effective antivirals in the case of other types of peptides. Lactoferricin and polyphemusin have ß structures stabilized by one and two internal disulfide bridges, respectively. These disulfide bridges have been shown to be crucial for the antiviral activity of the peptides (6, 120, 241) (Fig. 1; Tables 1 and 2).

Despite their diverse structures, many peptides possess analogous antiviral modes of action (119, 120), indicating that these peptides are able to interact with their targets, despite large structural differences. A possible explanation would lie in the observation that antimicrobial host defense peptides are known to adopt amphipathic conformations that are intrinsic to antibacterial activity and, we propose here, to antiviral activity. Interestingly, although the viral target of these peptides appears to vary, the demonstrated antiviral effects are quite similar.

Heparan sulfate is the most important glycosaminoglycan molecule with respect to viral attachment (166, 234); consequently, blocking of heparan sulfate can reduce the viral infection (222, 260). The importance of heparan sulfate for different viral infections varies considerably. It has been demonstrated that recombinant cells lacking heparan sulfate and chondroitin sulfate expression demonstrate an 80% and 60% reduction, respectively, in susceptibility to HSV infection (158). Enzymatic removal of cellular heparan sulfate and chondroitin sulfate has led to the observation that these proteoglycan molecules have minor influences on HIV attachment to host cells. However, it has been demonstrated that they are of major importance for HIV entry and replication (8). In contrast, only highly sulfated heparan participates in the entry of hepatitis C virus (14). Inhibitors of heparin sulfate biosynthesis, such as heparin, heparinase I treatment, and sodium chlorate, all demonstrate the ability to inhibit human cytomegalovirus (HCMV) infection in a dose-dependent manner (231). It has even been indicated that naked coxsackievirus B3 makes use of a specific modified heparan sulfate molecule for viral entry (274).

One might hypothesize that antimicrobial peptides that interact with heparan sulfate should be able to block many different viral infections. The large number of positively charged residues in antimicrobial peptides enables them to interact electrostatically with negatively charged cell surface molecules, including heparan sulfate. Human -defensin, LL-37, and magainin have all been shown to interact with different glycosaminoglycan molecules (116, 217, 218). Specific glycosaminoglycan binding domains have also been identified in bovine and human lactoferricin. These domains involve the sequence elements G₁RRRRS₆ and R₂₈KVR₃₁ in human lactoferricin (157) and the entire sequence of bovine lactoferricin (223). The sequence and structural diversity in these glycosaminoglycan binding peptides suggests that the critical factor driving interaction is how charged residues are presented in the secondary structure. Several lactoferricin analogues and synthetic -helical peptides have been made in an attempt to better understand these interactions. The results illustrate that the affinity for heparan sulfate is only partly correlated with the net charge of the peptides (119, 120). For example, analogues with Arg residues appear to promote a higher glycosaminoglycan affinity than comparable analogues substituted with Lys (67, 99, 119, 237).

It has been demonstrated that lactoferricin and a set of short -helical peptides are able to block HSV infection by binding to heparan sulfate in a way similar to that demonstrated by lactoferrin (5, 119, 120). This is supported by the fact that mixing the peptides with HSV prior to interaction did not increase antiviral activity (5). The peptides exhibited different antiviral effects for HSV type 1 (HSV-1) and HSV-2, an observation attributed to the combined effects of the amino acid content and the structures of the peptides (4, 119, 120, 122). Interestingly, differences in antiviral effects against HSV-1 and HSV-2 have also been reported for other polycationic and even polyanionic compounds (109, 138). These differences may reflect the viral specificity of particular receptor molecules and the differential ability of peptides to interact with the different viral receptors.

3-O-Heparan sulfate is a specific HSV entry receptor with structural similarities to the usual heparan sulfate, probably with increased affinity potential for cationic peptides. Peptide interaction with 3-O-heparan sulfate may result in interference or blocking of HSV glycoprotein D binding, resulting in specific inhibition of HSV entry. This might explain why several cationic peptides have demonstrated higher antiviral activity against HSV-1 than against HSV-2 (119, 120), since 3-O-heparan sulfate can serve as an entry receptor for HSV-1 and not for HSV-2 (261).

Bovine lactoferricin demonstrated an antiviral activity against human cytomegalovirus and human papillomavirus at the virus-cell interface (6, 55). Bovine lactoferricin also exhibited anti-HIV activity, and this might be related to heparan sulfate binding. Binding of HIV to the CD4 surface receptor is known to induce conformational changes in gp120 in the viral envelope, resulting in increased affinity for heparan sulfate. This finding implies that heparan sulfate is important at a later stage of the virus-cell attachment process (254).

(i) Blocking of cell-to-cell spread. The effect of antiviral peptides is also related to their ability to inhibit the spread of virus from one cell to a neighboring cell across tight junctions (cell-to-cell spread) or inhibition of giant cell (syncytium) formation. This is a property of the -helical alpha and gamma interferons, which reduce the cell-to-cell spread of HSV (167). Rabbit -defensin NP-1 has also been reported to inhibit both primary entry and cell-to-cell spread of HSV (225). Similar anti-HSV activity has been indicated for bovine lactoferricin (5, 119), while the polyphemusin analogue T22 and tachyplesin I have been demonstrated to inhibit syncytium formation in cocultures of persistently HIV type 1 (HIV-1)-infected cells (171, 177).

Blocking of viral entry by interaction with specific cellular receptors. Antimicrobial peptides might interact directly with specific viral receptors on the host cell (42, 240), influencing viral attachment, entry, or intracellular shuttling. The most obvious example of this is the known ability of the polyphemusin analogue T22 to bind to the chemokine receptor CXCR4, which serves as a coreceptor for HIV-1 entry into T cells (173, 243). Thus, it antagonizes that subgroup of HIV strains which use this chemokine receptor but not those that use CCR5 (263).

Recently a new HSV entry receptor was described (196); it is effectively blocked by binding of an -helical peptide in a coiled-coil formation (197). Whether this receptor is specific for HSV-1 or also allows HSV-2 entry is unknown; therefore, this cannot definitively explain differences in the antiviral effects of peptides on the two viruses. However, there is certain evidence that this receptor may be a potential target for several -helical cationic peptides (119). Similar coiled-coil domains (196) are found in fusion proteins such as the hemagglutinin of influenza virus and gp41 of HIV. Studies have illustrated that peptides mimicking these heptad repeat domains specifically interfere with membrane fusion and viral infection (115, 228).

Blocking of viral entry by interaction with viral glycoproteins. Antimicrobial peptide interactions with glycoproteins in the viral envelope have been proposed to influence the viral entry process. -Defensin (retrocyclin 2) interacts with the HSV-2 glycoprotein B with high affinity, thus protecting the cells from HSV-2 infection (270). The closely related retrocyclin-1 binds HIV gp120 with high affinity, as long as the envelope protein is glycosylated, probably resulting in an anti-HIV activity. This makes the -defensin the first antimicrobial peptide isolated from vertebrates with a lectin-like character (256). The polyphemusin analogue T22 has been demonstrated to inhibit fusion between the HIV envelope and the host cell membrane (177) through specific binding of the viral envelope protein gp120 and the T-cell surface protein CD4 (242).

Membrane or viral envelope interaction. (i) Viral envelope interaction. Antimicrobial peptides are known for their ability to interact with lipid membranes, resulting in destabilization, translocation, pore formation, or lysis (46, 226). This makes the viral envelope a potential target for direct interaction. Indolicidin causes a direct inactivation of the HIV-1 particle in a temperature-sensitive fashion, indicating a membrane-mediated antiviral mechanism (208). Dermaseptin also exerts an anti-HIV activity prior to viral entry by direct interaction with the viral particle, disturbing its organization and disrupting the viral membrane (153). In contrast, dermaseptin has no such direct effect on the HSV envelope. The anti-HSV activity of dermaseptin is proposed to result from blocking of viral entry by interaction with viral or cellular surface molecules involved in the attachment/adsorption/fusion phase of HSV (16). The antibacterial selectivity of dermaseptin depends in part on the lipid composition of the microbial membrane relative to that of the host cell (52). Similar requirements could also be hypothesized for the ability of antiviral peptides to interact with and destabilize viral membranes.

Human neutrophil peptide-1 (HNP-1) is an -defensin that neutralizes HSV-1 in a time-, temperature-, and pH-dependent manner. Neutralization of HSV-1 is also antagonized by serum or serum albumin (45). Preincubation of HSV-2 and HNP-1 prior to infection has been shown to reduce infection by >98%. In contrast, pretreatment of host cells with HNP-1 had no obvious effect on the anti-HSV activity. Although the peptide did not compete with viral envelope proteins for binding to cellular heparan sulfate, it still prevented viral entry (225).

A concentration-, time-, and temperature-dependent inactivation of vesicular stomatitis virus (VSV) is observed when the virus is incubated with tachyplesin-1 or its isopeptides prior to infection. Electron micrographs of the tachyplesin-1-treated VSV particle showed naked and damaged virions, confirming the direct effect of peptides on the viral envelope. Similar but weaker inactivation has been observed for influenza A virus (IAV) (type H1N1), while HSV-1, HSV-2, adenovirus 1, reovirus 2, and poliovirus 1 were resistant (174).

Neither cecropin, magainin, nor bovine or human lactoferricin possesses the ability to directly inactivate HSV when mixed together prior to infection (3, 5). Using electron microscopy, it has been demonstrated that bovine lactoferricin did not interact directly with the HSV particle, indicating that interactions with HSV glycoproteins do not occur (H. Jenssen, unpublished results).

(ii) Cellular membrane interaction. Host cell membranes are involved in several stages of viral interaction, and due to the ability of peptides to interact with and permeabilize membranes, this must be considered as a potential target. Similar permeabilization of the host cell membrane appears to occur (139), and the resulting alteration of host membranes could affect the efficiency of viral entry. An eight-residue cyclic DL--peptide has been shown to specifically prevent the lowering of pH in endocytic vesicles, thus arresting the entry of adenovirus particles by this pathway and abrogating the infection without having an apparent effect on host cell viability (102). This effect has been hypothesized to be a result of the peptide's membrane permeabilization properties. Similar effects could be anticipated for other membrane-acting antimicrobial peptides.

Intracellular targets and host cell stimulation. It is known that antimicrobial host defense peptides such as PR39 and LL-37 are able to cross lipid membranes, including the plasma and nuclear membranes of host cells, while others are constitutively located as precursors inside host cell vacuoles (5, 93, 139). Cellular internalization of these antimicrobial peptides can result in gene/protein stimulation, influencing host cell antiviral mechanisms (26), or might block viral gene/protein expression (255).

The effect of antimicrobial peptides has also been demonstrated to be crucially dependent on the experimental conditions. Salt may influence the structure of the peptides and their association with anionic cell molecules, thus affecting their antimicrobial activity (117); e.g., both the antibacterial and antifungal activities of cathelicidin and the antibacterial activity of ß-defensin are salt dependent (12, 230). Similarly, the antiviral mode of action of -defensin has been demonstrated to be serum dependent (37). This indicates that the peptide's action in vivo may be dependent on the physiological milieu at the site of infection. Transmission electron microscopy studies have revealed that human and bovine lactoferricins can translocate intracellularly (5). Bovine lactoferricin was able to enter both chondroitin sulfate- and heparan sulfate-deficient cells in an energy-independent manner (5). This mechanism was described earlier (66, 67, 239), and it appears that the arginine content of the peptides is an important factor, as it is a known feature of nuclear localization signals. As human lactoferricin has multiple Arg residues (195), this probably contributes to the shuttling of the peptide into the nucleus, where it can bind DNA. Consistent with this, the antiviral peptides LL-37 and indolicidin both can act as nuclear localization signals to translocate antisense nucleic acids (215).

Because of the known ability of peptides to interact with DNA (104, 123, 187, 232), one might speculate that they can directly influence viral nucleic acid synthesis, as shown for polyphemusin T22 and lactoferrin. Conversely, peptides are known to have immunomodulatory activities which include the upregulation of interferons and chemokines (24, 27, 37, 219), and thus peptides might exert their antiviral activities in part by stimulating the antiviral immune system of the host cell.

Direct effects of peptides on the intracellular steps of viral infection have been demonstrated. For example, cecropin A has been shown to inhibit Junin virus multiplication by reduction of its protein synthesis under conditions where the synthesis of host cell proteins remains unaffected (3). Melittin and cecropin A are also able to inhibit cell-associated production of HIV-1 by suppressing HIV-1 gene expression (255). Early steps in the HIV-1 replication cycle are inhibited by protegrin-1 (236), while HIV-1 integrase is effectively inhibited by indolicidin (134). Retrocyclin has been demonstrated to inhibit proviral DNA formation and protect immortalized and primary human CD4⁺ lymphocytes from in vitro HIV-1 infection (42). Transport of HSV-2 tegument protein VP16 to the cell nucleus and expression of ICP4 are effectively blocked in the presence of rabbit -defensin (NP-1) (225). In addition, the human -defensin-1 has demonstrated a direct inactivation of HIV-1 in the absence of serum, an effect that is abolished by the presence of serum. Conversely, in the presence of serum the peptide inhibits HIV-1 replication, partly by interfering with host cell protein kinase C signaling (37).

ANTIBACTERIAL ACTIVITY

Top Previous Next References

 
By far the best-studied class of cationic antimicrobial peptides are those with antibacterial activity (60). It is well understood that regardless of their actual target of action, all antibacterial cationic peptides must interact with the bacterial cytoplasmic membrane (92). The driving physical forces behind antibacterial activity have been defined in detail (see references 46, 91, and 92 for overviews) and include net positive charge (enhancing interaction with anionic lipids and other bacterial targets), hydrophobicity (required for membrane insertion and often driven by this process), and flexibility (permitting the peptide to transition from its solution conformation to its membrane-interacting conformation). Each of these characteristics can vary substantially over a particular range but are essential for the function of the peptides as antimicrobial agents and allow them to interact with bacterial membranes, which is critical to their exertion of antimicrobial effects.

The antibacterial activity of cationic peptides can also be modulated through alteration of the peptide's hydrophobicity or net charge. Studies have demonstrated that high levels of hydrophobicity can decrease selectivity between the desired bacterial targets and host cells (136, 275). Similarly, incorporation of charged residues above a certain maximum (varying with each peptide) does not lead to an increase in activity (46). Thus, this balance of charge and hydrophobicity can be delicate and must be empirically determined for each series of peptides.

Consequently, the inclusion of a particular peptide into a structural group does not give an indication as to its mode of action or its spectrum of activity. In fact, some peptides with very similar secondary structures have quite opposite characteristics with respect to antibacterial activity (Table 3). The -helical melittin from bees, which is thought to perforate the membranes of both prokaryotic and eukaryotic organisms, falls within the -helical structural class. Conversely, the -helical toad peptide buforin translocates into cells and acts on macromolecular synthesis. Studies of -helical analogues of a cecropin-melittin hybrid peptide have revealed that even peptides that have similar secondary structures and minimal differences in the primary sequence can possess quite different antibacterial activities (64). Indeed, different peptides may be membrane permeabilizing at their minimal effective concentrations or at concentrations well above or well below these concentrations. Nonetheless, antibacterial peptides seem largely able to effect their antimicrobial activity because of their amphipathicity or amphiphilicity and because of the presence of regions within the folded structure with high concentrations of positively charged residues (202).

Membrane-permeabilizing peptides. Several different models have been proposed to explain how, following initial attachment, antibacterial peptides insert into the bacterial membrane to form transmembrane pores which result in membrane permeabilization (Fig. 2). The amphipathic nature of antimicrobial peptides is a key characteristic for this process, as hydrophobic regions are necessary to interact directly with the lipid components of the membrane, while hydrophilic regions either interact with the phospholipid head groups or face the lumen of the pore. Generally, these models can explain the pore-forming ability of -helical antibacterial peptides; however, the mechanisms utilized by ß-sheet peptides such as defensins have not been as well studied. While ß-sheet antimicrobial peptides can adopt amphipathic folds, there is little experimental evidence to indicate which of the following models is applicable to defensins.

View larger version (66K): [in this window] [in a new window]   FIG. 2. Mechanisms of action of antibacterial peptides. The bacterial membrane is represented as a yellow lipid bilayer with the peptides shown as cylinders, where the hydrophilic regions are colored red and the hydrophobic regions are blue. Cell wall-associated peptidoglycan molecules are depicted as purple cylinders. Models to explain mechanisms of membrane permeabilization are indicated (A to D). In the "aggregate" model (A), peptides reorient to span the membrane as an aggregate with micelle-like complexes of peptides and lipids, but without adopting any particular orientation. The "toroidal pore" model (B) proposes that peptides insert perpendicular to the plane of the bilayer, with the hydrophilic regions of the peptides associating with the phospholipid head groups while the hydrophobic regions associate with the lipid core. In this process, the membrane also curves inward such that the bilayer also lines the pore. In the "barrel-stave" model (C), the peptides insert in a perpendicular orientation to the plane of the bilayer, forming the "staves" in a "barrel"-shaped cluster, with the hydrophilic regions of the peptides facing the lumen of the pore and the hydrophobic regions interacting with the lipid bilayer. The "carpet" model (D) proposes that peptides aggregate parallel to the lipid bilayer, coating local areas in a "carpet"-like fashion. At a given threshold concentration, this is thought to result in a detergent-like activity, causing formation of micelles and membrane pores. The mechanisms of action of peptides which do not act by permeabilizing the bacterial membrane are depicted in panels E to I. The antimicrobial peptides buforin II, pleurocidin, and dermaseptin have all been shown to inhibit DNA and RNA synthesis at their MICs without destabilizing the membrane (E). Protein synthesis is another macromolecular target for antibacterial peptides such as indolicidin and PR-39, which have been shown to decrease the rate of protein synthesis in target bacterial cells (F). Several antibacterial peptides have been shown to act on other intracellular target processes, such as enzymatic activity. The ATPase activity of DnaK, an enzyme involved in chaperone-assisted protein folding, is targeted by pyrrhocidin (G), while inhibition of enzymes involved in the modification of aminoglycosides has also been demonstrated (H). Antimicrobial peptides can also target the formation of structural components, such as the cell wall (I). Lantibiotics such as nisin and mersacidin can bind to and inhibit, respectively, the transglycosylation of lipid II, which is necessary for the synthesis of peptidoglycan.

In contrast to the barrel-stave and toroidal pore models, the carpet model proposes that aggregates of peptide align parallel to the lipid bilayer, coating local areas in a carpet-like fashion (200) (Fig. 2D). At sufficiently high concentration, this is thought to have a detergent-like activity (causing patches of the membrane to break up into micelles), causing local disturbances in membrane stability which can lead to the formation of holes in the membrane. Many cationic antimicrobial peptides will do this at high enough concentrations due to their amphipathic character; however, there is very limited evidence demonstrating that most peptides cause membrane dissolution at the minimal effective concentrations in vivo (or, in many studied cases, in vitro; for example, Zhang et al. [279] examined nine representative peptides and demonstrated that most of them were able to cause membrane flip-flop at far lower concentrations than those at which they could cause calcein release). Based on mechanistic studies, however, this mechanism has been proposed for ovispirin, a rationally designed antibacterial peptide based on the sheep cathelicidin Smap29 (264). It is important to recognize that each of these models may have validity under different circumstances, and examination of a broad range of peptides with different sizes and structures has indicated that they leave quite different signatures of membrane interaction (280), while differential scanning calorimetry studies on LL-37 (98) and polyphemusin (203) have shown opposite effects on membrane curvature.

Peptides that do not act by membrane permeabilization. All peptides must interact with the cytoplasmic membrane, if only to get to their site of action. It is now well established that several antimicrobial peptides do not cause membrane permeabilization at the minimal effective concentration yet still result in bacterial death. A growing number of peptides have been shown to translocate across the membrane and accumulate intracellularly, where they target a variety of essential cellular processes to mediate cell killing. Novel modes of action that have recently been demonstrated include inhibition of nucleic acid synthesis, protein synthesis, enzymatic activity, and cell wall synthesis (28) (Fig. 2).

The frog antimicrobial peptide buforin II translocates across the bacterial membrane without causing permeabilization and binds to both DNA and RNA within the cytoplasm of E. coli (187). Similarly, -helical peptides such as derivatives of pleurocidin, a fish-derived antimicrobial peptide, and dermaseptin, isolated from frog skin, cause inhibition of DNA and RNA synthesis at their MICs without destabilizing the membrane of E. coli cells (193, 238) (Fig. 2E). Inhibition of nucleic acid synthesis has also been demonstrated for antimicrobial peptides from different structural classes, such as the ß-sheet human defensin, HNP-1 (147), and the extended-structure bovine peptide indolicidin (238). Additionally, some of these peptides have been shown to interfere with protein synthesis. Pleurocidin and dermaseptin can block tritiated leucine uptake in E. coli, and PR-39- and indolicidin-treated cells also exhibit reduced rates of protein synthesis (22, 65, 193, 238) (Fig. 2F).

Inhibition of cellular enzymatic activity by proline-rich insect antimicrobial peptides has also been observed. Pyrrhocidin enters the target cell and binds to DnaK, a heat shock protein that is involved in chaperone-assisted protein folding. Specifically, the peptide inhibits the ATPase activity of DnaK, preventing protein folding, which results in the accumulation of misfolded proteins and cell death (133, 184) (Fig. 2G).

Antimicrobial peptides can also target the formation of structural components, such as the cell wall (Fig. 2I). The bacterially produced lantibiotic mersacidin interferes with transglycosylation of lipid II, a necessary step in the synthesis of peptidoglycan (29). Nisin, another lantibiotic, can also bind to lipid II, thus inhibiting cell wall synthesis in addition to its pore-forming activity. Interestingly, this is the same biosynthetic process that is targeted by the antibiotic vancomycin; however, mersacidin and nisin are thought to act by interacting with distinct molecular moieties within lipid II, explaining why these peptides are still active against vancomycin-resistant bacteria (31, 135).

It is likely that the mode of action of individual peptides may vary according to the particular bacterial target cell, the concentration at which they are assayed, and the physical properties of the interacting membrane. It is also likely that in the context of infection, antimicrobial peptides may use more than one mechanism of action, such as destabilization of the cell membrane combined with inhibition of one or more intracellular targets. We recently proposed a "multitarget" mechanism (201), which hypothesizes that highly charged cationic peptides can bind to anionic molecules within the cytoplasm, such as nucleic acids or enzymes with ionic surfaces, thus interfering with processes in which these molecules are involved. An example is that of aminoglycoside-modifying enzymes, which contain an anionic binding pocket. It was recently demonstrated that cationic peptides of various structural classes can bind and specifically inhibit the activity of these enzymes (20) (Fig. 2H). This high degree of complexity of the mechanism is almost certainly the cause of the observation that it is extremely difficult to select cationic-peptide-resistant mutants (235, 277).

ANTIFUNGAL ACTIVITY

Top Previous Next References

 
Our knowledge of antifungal peptides has accelerated in recent years. Their mode of action was first described as involving either fungal cell lysis or interference with fungal cell wall synthesis (51). However, as the numbers of known antifungal peptides increase, new modes of action are being identified (Table 4). It is intriguing to note that peptides with primarily antifungal activity, such as many of those isolated from plants, tend to be relatively rich in polar and neutral amino acids, suggesting a unique structure-activity relationship (155).

Modification of ineffective antimicrobial peptides has revealed that relatively modest changes often result in antifungal activity. For example, conjugation of undecanoic acid or palmitic acid to magainin resulted in analogue peptides that had gained potent activity against both yeast and opportunistic fungal infections (9). The fusions of parts of magainin 2 and cecropin A to form the hybrid peptide P18 resulted in a peptide with potent fungicidal activity against pathogenic Candida albicans, Trichosporon beigelii, Aspergillus flavus, and Fusarium oxyspovrum (140). Studies with the 18-residue pig peptide protegrin, which has both antibacterial and antifungal activities, demonstrated that the antibacterial activity could be retained in a 12-residue deletion peptide but only two residues could be deleted for the potent antifungal properties of this peptide to be retained (41). Other studies with synthetic 12-mers based on the bovine peptide bactenecin indicate that different substitutions were required for optimizing antibacterial and antifungal properties (100).

Although no conserved sequences are evident for the antifungal peptides, several have been demonstrated to possess specific biochemical characteristics, such as chitin (69, 105) or heparin (7, 223) binding abilities. Structure-activity relationship studies on three synthetic bovine lactoferricin (amino acids 17 to 30)-derived peptides revealed a significant positive correlation between the pI values of peptides and their candidicidal activity (181). Another study also showed a direct correlation between the ability of peptides to form complexes with lipid mixtures and their antifungal activity (152).

The mechanism of action of certain antifungal peptides is still a matter of debate. The formation of reactive oxygen species has been suggested to be the crucial step in the fungicidal mechanism of a number of antimicrobial peptides (179), including histatin 5 and lactoferrin-derived peptides (97, 154). Conversely, Veerman et al. have concluded that reactive oxygen species do not play a role in the histatin 5-mediated killing of C. albicans (252). Helmerhorst et al. (96) found that exposure to the cationic peptide histatin 5 caused a depletion of mitochondria in C. albicans. Histatins from saliva of humans and some other higher primates bind to a receptor on the fungal cell membrane and enter the cytoplasm, where they target the mitochondrion (124). Thus, it is reasonable to hypothesize that antimicrobial peptides may interact with mitochondria in a manner very similar to some of their actions on bacteria, as suggested by the studies of Helmerhorst et al. (96).

The glycine-rich peptide tenecin-3, both native and recombinant, has been demonstrated to possess candicidal activity (126). The peptide is able to quickly enter the cytoplasm of C. albicans in an energy-dependent mechanism influenced by the metabolic status of the target cells and the ionic environment (125). It does not result in membrane permeabilization or depolarization (125), although pegylated tenecin-3 peptide can result in K⁺ leakage from C. albicans (146), possibly explaining its improved antifungal activity. The antifungal mechanism of native tenecin-3 is unknown, but the loss of cell viability occurs after the peptide enters the host cell cytoplasm (125).

Pn-AMP1, is a small cysteine-rich plant peptide that causes depolymerization of the actin cytoskeleton in Saccharomyces cerevisiae and C. albicans. It was suggested that cell wall proteins worked as determinants of Pn-AMP1 resistance, while its ability to induce actin depolymerization was important for its antifungal activity (131).

ANTIPARASITIC ACTIVITY

Top Previous Next References

 
Magainin 2 was one of the first antimicrobial host defense peptides demonstrated to display antiprotozoan activity, leading to swelling and eventual bursting of Paramecium caudatum (272). More recently an acylated synthetic antimicrobial peptide, Oct-CA(1-7)-M(2-9), has been shown to be both safe and effective for treating naturally acquired canine leishmaniasis (2), which is caused by the parasite Leishmania, which is also an important cause of morbidity and mortality in humans (48).

The antinematodal effect of the porcine cathelicidin PMAP-23 has been demonstrated against both the eggs and worms of Caenorhabditis elegans. Studies have indicated that this effect is exerted through disruption of the cell membrane via pore formation or via direct interaction with the lipid bilayers (190), resembling the antifungal mode of action for PMAP-23 (143).

Several antimicrobial peptides possess an antiprotozoan mode of action that indicates parallels with their antibacterial, antiviral, or antifungal modes of action. Analogues of mussel defensins have been demonstrated to efficiently kill Leishmania major and Trypanosoma brucei in a temperature-, time-, and dose-dependent manner. These peptides were found to interact with the external epithelium of T. brucei. However, structure/activity relationship studies indicated that the antiprotozoan and antiviral activities were mediated by different mechanisms (209). Analogous studies on magainin 2 analogues revealed that short stretches of hydrophobic amino acids were important for leishmanicidal activity (82). It seems likely that antiprotozoan activity may be dependent on peptide motifs fundamentally different from those required for bacterial, viral, and fungal activities.

DEVELOPMENT OF ANTIMICROBIAL PEPTIDES FOR CLINICAL APPLICATIONS

Top Previous Next References

 
As the problem of emergence of bacterial resistance to current antibiotic drugs continues to grow, there has been considerable interest in the development of antimicrobial peptides as a novel therapeutic approach to treat infections. To date, several antimicrobial peptides have been developed and entered into clinical trials, and there are also peptides that are currently in the preclinical development stage, which are discussed later in this section. Antimicrobial-peptide-based therapies are attractive candidates as alternative antibiotic treatments, since they offer several potential advantages over currently used classes of drugs. First, they represent a naturally occurring means of combating p