Thiotemplate Mechanism Of Peptide Antibiotic Synthesis (Molecular Biology)

Peptide antibiotics are unusual in that they contain many unusual and modified amino acid residues, even D-amino acid residues, and often have cyclic polypeptide chains. Although some are synthesized on ribosomes, in the usual manner of protein biosynthesis, and are then post-translationally modified, many have unusual, nonribosomal mechanisms of synthesis. The thiotemplate mechanism refers to this type of mechanism of forming peptide bonds sequentially to generate a polypeptide precursor of peptide antibiotics. The name was coined when thioester intermediates and the presence of 4′-phosphopantetheine, derived from coenzyme A, were discovered in an enzyme system catalyzing the synthesis of the cyclic decapeptide gramicidin S (1). It was suggested that amino acid residues were polymerized sequentially on enzyme subunits as thioesters attached to the 4′-phosphopantetheine swinging arm of a carrier protein. It transformed the fatty acid elongation cycle proposed by Lynen (2) to enzymatic peptide formation and introduced the concept of a protein template for assembly of a peptide sequence (3).

The thioester type of assembly mechanism has been established for a wide variety of peptides, including tyrocidine, bacitracin, polymyxin, enniatin, beauvericin, actinomycin, alamethicin, ferrichrome, cyclosporin, and the penicillin precursor peptide d-(L-a-aminoadipoyl)-L-cysteinyl-D-valine. It soon became apparent that it was the principal method of forming peptide bonds in the absence of ribosomes, and it has been shown to operate for at least 25 steps, polymerizing that number of amino acids specifically (4).


When it became evident that the protein template enzymes were unusually large structures, with masses often exceeding 500 kDa, Kurahashi introduced the term multienzyme thiotemplate mechanism (5). It was later pointed out, however, that the principle of the 4′-phosphopantetheine swinging arm did not account for all of the experimental observations, including (i) the accumulation of all intermediates preceding a certain step when the process was interrupted there, (ii) the loss of thioester formation, with retainment of adenylate activation, and (iii) constraints on the access of a single cofactor on an integrated carrier domain to 11 amino acid intermediates in a system like cyclosporin synthetase.

These issues were resolved after the amino acid sequences of several peptide synthetases were determined from the sequences of their genes (6). This revealed that the peptide synthetases had modular structures composed of (i) adenylate-forming activation domains, (ii) carrier proteins containing sites for attachment of the respective cofactor, (iii) condensation domains, and (iv) a variety of modification functions responsible for epimerization, methylation, cyclization and hydrolytic release of the precursor peptide. Most striking was the absence of conserved cysteine residues that had been implied by the thioester concept. It became obvious that each activation domain was associated with a carrier protein homologous to the acyl carrier protein of the fatty acid synthesis systems, and the subsequent addition of 4′-phosphopantetheine to the carrier proteins has been demonstrated by mass spectrometry. These findings have led to a further refinement of the process, which is now called the multiple carrier thiotemplate mechanism (7).

1. Multiple Carrier Thiotemplate Mechanism

To illustrate the multiple carrier thiotemplate mechanism, a scheme for the formation of a tripeptide is illustrated in Figure 1. Each amino acid, A1, is adenylated and activated by a specific adenylation domain of the enzyme and then transferred to the thiol group of the pantetheine cofactor residing on its respective carrier domain, Ei-SH, as a thioester, Ei-S-Ai. Peptide bond formation is initiated with the amino acid that will be C-terminal, A , becoming linked to the next amino acid, A , Elongation proceeds similarly and sequentially, adding a third and further amino acid thioesters to the peptide:

tmp10D-20_thumb[1]tmp10D-21_thumb[1]

In this way, a polypeptide chain is synthesized starting from the carboxyl terminus. At each step, the assembly process is specific for the amino acid to be incorporated and for the appropriate enantiomer, either D or L. Hydroxy acids can also be incorporated, and the corresponding ester bonds are formed similarly, using acyl intermediates. The process thus resembles in principle the ribosomal mechanism, in that there is transfer of an activated peptidyl intermediate to an aminoacyl intermediate (8). Release of the complete peptide chain occurs on the appropriate termination domain, either as the linear chain, which is produced by hydrolysis or aminolysis, or as the cyclic peptide or as peptidolactone.

Figure 1. Illustration of the multiple carrier thiotemplate mechanism, as elucidated for synthesis of the d-(L-a-aminoadipoyl)-L-cysteinyl-D-valine tripeptide b-lactam precursor (11). Each of the three amino acids is activated at a specific adenylation domain, EA. Each adenylate domain is linked to a specific carrier domain, Pi. The aminoacyl moietie a1 of the Ai-AMP adenylates are transferred to the respective thiol groups of the 4′-phosphopantetheine cofactor, illustra by the thick zigzag line. Peptide bond formation occurs, in analogy to the ribosomal cycle, from amino acyl (A) and pept (P) sites on specific condensation domains. In the case of the initiation reaction, the P-site is called the initiation site, or ] site. The tripeptide formed, A3-A2-A\ is thioester bound and may be transferred for further elongation or released by a thiolesterase, as in penicillin formation.

Illustration of the multiple carrier thiotemplate mechanism, as elucidated for synthesis of the d-(L-a-aminoadipoyl)-L-cysteinyl-D-valine tripeptide b-lactam precursor (11). Each of the three amino acids is activated at a specific adenylation domain, EA. Each adenylate domain is linked to a specific carrier domain, Pi. The aminoacyl moietie a1 of the Ai-AMP adenylates are transferred to the respective thiol groups of the 4'-phosphopantetheine cofactor, illustra by the thick zigzag line. Peptide bond formation occurs, in analogy to the ribosomal cycle, from amino acyl (A) and pept (P) sites on specific condensation domains. In the case of the initiation reaction, the P-site is called the initiation site, or ] site. The tripeptide formed, A3-A2-A\ is thioester bound and may be transferred for further elongation or released by a thiolesterase, as in penicillin formation.

The reactions linking the successive amino acids and analogues are thought to be catalyzed by the condensation domains; they are essentially irreversible. In addition, catalytic sites may be involved in the modification of the thioester intermediates, such as their #-methylation or epimerization and, especially in the case of polyketide biosynthesis, in the reduction and dehydration of the respective keto intermediate.

The key component of the thiotemplate mechanism is the acyl-, aminoacyl-, or peptidyl-carrier protein or domain, to which the respective intermediate remains attached covalently during the elongation cycle. Carrier domains can be identified readily by the conserved elements of their protein structure, including a cavity formed by three a-helices, an additional a-helix, and a variable loop containing the 4′-phosphopantetheine cofactor binding site (9). Addition of the cofactor to the carrier domain is catalyzed by a family of 4′-phosphopantetheine transferases, using coenzyme A (10). Condensation processes require at least one carrier protein, and stepwise processes involve the intra- or intermolecular transfer of acyl intermediates between two carrier protein proteins. The largest known thiotemplates contain 11 carrier domains in peptide biosynthesis, for cyclosporin, and six carrier domains in polyketide biosynthesis, for rapamycin.

The thiotemplate mechanism contrasts with the mechanism of formation of other small peptides, such as glutathione or peptidoglycan components, which are readily distinguished by (i) their use of phosphate to activate the amino acids, rather than adenylate, (ii) the absence of covalently attached intermediates, and (iii) the activation of peptide carboxyl groups, rather than amino acid carboxyls.

Next post:

Previous post: