Peptide Antibiotic Biosynthesis (Molecular Biology)

Peptide antibiotics include a variety of peptide structures with diverse biological properties (1). Historically, the first antibacterial compounds identified were peptides (penicillin, gramicidin, tyrocidine). New such peptides have been characterized steadily, including actinomycin (antitumor); cyclosporin (immunosuppressant); destruxin (insecticide); ferrichrome (siderophore); pepstatin, antipain, and bestatin (proteinase inhibitors); microcystin (phosphatase inhibitor); and defensin and magainin (anti-infectives).

Their biosynthesis may occur either (i) on ribosomes by direct transcription, translation, and posttranslational processing of peptide antibiotic genes (see Protein biosynthesis)—for example, nisin—or (ii) without ribosomes but on a multienzyme system, using the thiotemplate mechanism. The structural complexity of peptide antibiotics ranges from modified amino acids to polypeptide chains of up to 50 amino acid residues (2). Ribosomal systems can incorporate only the L isomer of the 20 normal amino acids, plus selenocysteine, while there are no such limitations to the nonribosomal systems, which can use hydroxy acids, fatty acids, and amines. Nevertheless, peptides of ribosomal origin are known to include D amino acid residues, cyclic side chains, and disulfide bonds. All nonribosomal peptides also originate from linear precursors, but a variety of types of cyclization are known. Identification of the biosynthetic origin is crucial for subsequent genetic analysis, and the structural properties of compounds arising from the two types of pathway are compared in Table 1.

Table 1. Structural Features of Peptide Antibiotics of Ribosomal and Nonribosomal Origin

1. Ribosomal Biosynthesis

Peptide antibiotics that are synthesized initially on ribosomes comprise a rapidly expanding field of research (3, 4). The compounds synthesized in this way range from linear polypeptide chains to compact multicyclic peptides with several disulfide bonds or thioether linkages (lantibiotics). Their genes are usually isolated by reverse genetics, and they usually reside within gene clusters, together with genetic information for their post-translational modification and export, and for resistance of the host to their actions. Genes for typical animal peptides, like defensins, are present in multiple copies with minor differences in sequence (5). The signals for their targeting and processing are critical for their functions in innate immunity. A peptide from the venom from the spider Angelopsis aperta is activated by epimerization of a specific residue (6).

Internal cyclizations other than disulfide bond formation have been restricted to bacterial sources, and the enzymes involved in formation of thiazolidines (microcin B) and lantibiotic ethers (dehydration of serine and threonine residues to dehydroalanine and dihydrobutryine, respectively, followed by addition of the thiol group of cysteine) are being characterized (7). The post-translational modification of these antibiotics takes place on a membrane-attached multienzyme complex, which also facilitates their export and their role in intercellular communication.

2. Synthesis by Peptide Synthetases

The biosynthesis of peptide antibiotics by nonribosomal systems, on peptide synthetases, requires a considerable amount of information, and these multienzyme complexes are among the largest protein structures known, with masses up to 1700 kDa, corresponding to a 45.5 kbp open reading frame, to synthesize cyclosporin (see Gene Structure). Synthesis of each amino acid residue of such a peptide antibiotic requires a minimum of three catalytic domains: (i) the activating adenylate domain, (ii) the carrier protein, and (iii) the condensing domain, which collectively are referred to as a module (see Thiotemplate Mechanism Of Peptide Antibiotic Synthesis). The nonribosomal code for the amino acid sequence of the peptide is determined by the substrate specificity of the activating domain. Selection and activation of the appropriate amino acid is followed by its attachment as a thioester to the 4′-phosphopantetheine prosthetic group attached to the adjacent carrier domain (8); it is then linked covalently to the next amino acid, on the adjacent module. The thioester intermediates may be subjected to various modifications, such as epimerization, N-methylation, and hydroxylation, catalyzed by additional domains introduced into the module structure. The intermediates in the reaction are covalently attached to the peptide synthetase, and only the final product is released. Examples of known peptide synthetases are compiled in Table 2.

Table 2. Nonribosomal Peptide Synthetase Systems

^a The abbreviations used are: P, peptide; C, cyclopeptide; L, lactone; E, ester; R, acyl; M, modified. The structural types are defined by the number of amino-, imino-, or hydroxy acids in the precursor chain. The ring sizes of cyclic structures are indicated in the number following C, L, or E, defining the type of ring closure. The abbreviations used for unusual amino acids and other compounds are listed in the abbreviations footnote on the first page.

Current applications of such enzyme systems are the enzymatic synthesis of peptide analogues, taking advantage of the relatively low stringency of the nonribosomal code—that is, the specificities of the peptide synthetases (9). Efforts are being made to alter the modular construction of peptide synthetases so as to generate new enzyme systems capable of synthesizing novel peptides.