Antimicrobial Peptides – The Defence Never Rests

Kenneth M. Huttner    Division of Neonatology, Department of Pediatrics, Harvard Medical School and the Mass-General Hospital for Children, Boston, MA 02114 U.S.

1 ANTIMICROBIAL PEPTIDES

1.1 Structures

The two salient features that identify an antimicrobial peptide are an ability to inactivate microorganisms, tested either in vitro or in vivo, and a size less than 100 amino acids. Given these rather broad definitions, over 500 molecules derived from either plant or animal species are now categorised as antimicrobial peptides [1–3]. As a general rule, they are derived from propeptides following constitutive or inducible processing. The active peptides vary in structure from disulfide-bridged circular peptides to amphipathic α-helical sequences enriched for a single amino acid (Fig. 1). A common feature is the formation of a tertiary structure permissive for interaction with the predominantly anionic, cholesterol-poor microbial membrane [4].

Many excellent reviews are available on the numerous classes of antimicrobial peptides, describing in a wealth of detail their gene, mRNA and peptide structures [5,6]. Here we will outline briefly the principle classes of antimicrobial peptides isolated from mammalian tissues, with particular emphasis on those from H. sapiens.

In the 1980’s, investigators at UCLA published seminal work characterising neutrophil defensins, the most abundant class of mammalian antimicrobial peptides [5]. As a general rule, mammalian defensins are cationic, arginine-rich, cysteine-containing peptides 29–47 amino acids in length that are derived from larger, precursor molecules. Most commonly, these precursor molecules are prepropeptides containing a signalling pre- sequence, a neutralizing propeptide and the carboxy-terminal defensin that is liberated following multiple processing steps. Nearly all mammalian defensins contain six cysteine molecules participating in three disulfide bonds. They can be sorted broadly into three groups based on cysteine spacing and disulfide-bond alignment (Fig. 1).

The α-defensins are generally shorter in length and contain disulfide bridges in a {C1-C6, C2-C4, C3-C5} pattern, while the (β-defensins are more variable in length at both termini and contain disulfide bridges in a {C1-C5, C2-C4, C3-C6} pattern. The core tertiary structure of α- and β-defensins is considered to be a triple-stranded antiparallel [β-sheet [7]. The third group of defensins, θ-defensins, has been identified only in rhesus macaques [8], and is unique both in structure (circular) and in mechanism of post-translational assembly (intermolecular disulfide bridging).

As an interesting aside, it was recognised recently that the CXC motif present in the N-terminal of defensins is analogous to the CXC motif present in a class of chemokines [9]. Specific CXC chemokines demonstrate a pattern of broad-spectrum antimicrobial activity similar to that of the defensins [9], perhaps blurring the distinction between antimicrobial and signalling molecules.

The second main class of mammalian antimicrobial peptides is the cathelicidins, all of which are cleaved from precursor proteins containing the highly conserved propeptide sequence termed cathelin [6]. Interestingly, the cathelin propeptide was discovered independently as a porcine inhibitor of cathepsin L [10,11]. Cathelicidin genes consist of 4 exons, with exons 1-3 encoding the tissue-targeting cathelin portion of the molecule, and exon 4 encoding the active peptide. The active cathelicidin peptides vary widely in their primary sequence, and the cathelicidin gene family varies markedly in size among species. To date, only a single human cathelicidin gene has been identified, hCAP18. hCAP18 refers to the 18 kd cathelicidin prepropeptide, while the processed, active moiety is termed LL-37 (N-terminal leu-leu with length 37 a.a.) (Fig. 1).

Peptides of the defensin and cathelicidin families follow the dual paradigm of being processed from a precursor propeptide and having an antimicrobial activity linked to their net cationic charge. However, nature in its abundant diversity is likely to have evolved multiple paradigms for providing molecular protection of the epithelial barrier. The search for additional, novel classes of antimicrobial peptides has required novel insights and novel laboratory approaches. Among the unique peptide classes under investigation are anionic peptides and histone fragments.

The former are zinc-binding small molecules detected in pulmonary tissues of both sheep and humans [12]. In both species, these molecules and their precursors are constitutively expressed in airway tissues [13]. It is interesting to speculate whether they contribute to the proven effectiveness of zinc supplements in ameliorating pulmonary infections.

Many investigators have reported the HPLC purification of histone fragments with antimicrobial activity, and for years this was thought to be of no real consequence. However, it is now better appreciated that cellular apoptosis may be a defence response to microbial invasion. Therefore, it is a reasonable hypothesis that the release of histone fragments from the apoptotic cell chromatin is a highly effective strategy for augmenting host defence [14,15].

1.2 Cellular origin

Each species examined has a unique pattern of antimicrobial peptide expression, perhaps reflective of its unique life cycle and environment. However, a feature common to all species is the presence of abundant antimicrobial peptides in circulating granulocytes (e.g., neutrophils [5] and eosinophils [16]). It has been estimated that the antimicrobial α-defensins stored in human neutrophil granules (termed HNP’s) may constitute up to 50% of the total neutrophil protein [5]. A subset of these granule antimicrobial peptides can be detected in epithelial tissues, with the most prominent examples being the epidermal expression of human cathelicidin LL-37 and the orthologous mouse cathelicidin CRAMP [11,17]. Additionally, cells of both the T and B cell lineages, capable of infiltrating epithelial tissues, demonstrate transcription, translation and release of both defensins and cathelicidins [15].

Defensins and cathelicidins are the principle antimicrobial peptides expressed in the epithelium of every major organ system examined. In human tissues this includes the skin, ear canal, salivary glands, conjunctiva, breast tissue, GI tract from oesophagus to colon, pulmonary tract, and genitourinary tract [6,11,18–20]. In addition to their expression in the epithelial lining cells, antimicrobial peptides may be released locally from tissue monocytes encountering microbes or their products [21]. Their presence in circulating phagocytic cells as well as at sites of host-environment interface has lead investigators to hypothesise that they are the microbicidal effector molecules of the innate immune system. Recent evidence described below (section 1.8–1.10, and section 2) not only is supportive of this role in innate immunity, but also is suggestive of a more central role in regulating host immunity.

1.3 Mechanisms of microbicidal action

In general, defensins and cathelicidins are bactericidal in vitro against both Gram-positive and Gram-negative bacteria in assays using simplified, low ionic strength media. Subsets of these peptides demonstrate antiviral, antifungal and antiprotozoal activity as well. A working model for their mechanism of activity includes an initial interaction of the cationic moiety with the anionic bacterial membrane phospholipids, subsequent weakening of the microbial membrane integrity and a terminal disruption of membrane function [4,22,23]. This model is consistent with the repeated observations that: (1) elevation of salt or divalent cation concentration in the media may shield the charges on the microbial surface and reduce antimicrobial peptide activity; and (2) bacteria may acquire resistance to microbial activity through alterations in their surface lipid properties.

As we attempt to develop a unifying model for the antimicrobial peptide mechanism of action, we should recognise that a number of independent...