Five enzyme families can assemble short peptides without the ribosome: non-ribosomal peptide synthetases, ATP-grasp enzymes, α-amino acid ester acyltransferases, β-lactam acylases, and cyanophycinases. They differ in activation chemistry, ATP cost, substrate scope, and structural characterization.…
Five families of enzymes can assemble short peptides without the ribosome: non-ribosomal peptide synthetases NRPSs , ATP-grasp enzymes, α-amino acid ester acyltransferases AETs , β-lactam acylases, and cyanophycinases. They are not interchangeable. The deciding differences are how each enzyme activates the carboxyl group that will form the new amide bond, whether that activation costs ATP, which substrates are tolerated, and how much structural information exists for engineering. This article works through the families as classified in a 2019 review by Wang and colleagues in Biomolecules 9 11 : 733, focusing on what a researcher or manufacturer can actually do with each route.
NRPSs are the most versatile of the five families. They are universal in bacteria and fungi and account for a large share of therapeutically important microbial products, including penicillin, bleomycin, and cyclosporine. That list carries a caution: penicillin and bleomycin are not peptides in the strict biochemical sense, since a β-lactam ring or a glycosylated scaffold is not a plain peptide chain. The review groups them as NRPS products because their biosynthetic logic is the same modular, thioester-based assembly used for peptides.
The review divides NRP-synthesizing biocatalysts into ATP-dependent and ATP-independent classes by how the substrate is activated. ATP-dependent enzymes activate their substrate as an aminoacyl-adenosine monophosphate, the strategy used by tRNA-dependent ligases and by the adenylation domains of NRPSs. ATP-independent enzymes instead activate through aminoacyl phosphate, as transacylases do. The distinction matters for cost: every aminoacyl-AMP activation consumes ATP, while the phosphate route does not.
A typical NRPS is a large multidomain protein containing A, C, Te, and PCP domains. The adenylation A domain selects and activates the amino acid, the peptidyl carrier protein PCP carries the growing chain as a thioester, the condensation C domain forms the amide bond between the upstream thioester and the downstream amino group, and the thioesterase Te domain releases the finished product. The review illustrates this arrangement with the structure of a typical NRPS, PDB ID 2VSQ, and shows amide bond formation schematically, without detailing each catalytic step.
For practical use, NRPSs are the route of choice when the target molecule is an existing NRPS product or a close analog. Production is typically whole-cell fermentation in the native or a heterologous host, because the enzymes are too large and too modular for convenient in vitro reconstitution. The trade-off is that engineering an NRPS to make a new product is slow; swapping domains or modules often breaks the assembly line.
ATP-grasp enzymes, also called ATP-dependent carboxylate-amine ligases, take a different activation route. They phosphorylate the carboxylate of one substrate using ATP, generating an acylphosphate intermediate, and the amine of the second substrate then displaces the phosphate to form the amide bond. The family is distributed across primary metabolism, including de novo purine biosynthesis, where glycinamide ribonucleotide synthetase carries out one of the early steps. This enzyme is the typical ATP-grasp example in the review, shown with PDB ID 2IP4.
Notable family members include biotin carboxylase, D-alanine-D-alanine ligase Ddl , and glutathione synthetase. These names show the functional range: Ddl builds the D-Ala-D-Ala dipeptide used in bacterial peptidoglycan, and glutathione synthetase completes glutathione. ATP-grasp enzymes share a characteristic architecture of three conserved domains and a nonclassical ATP-binding fold that encloses ATP, and most require an Mg2+ ion coordinated by ATP in the active site.
Peptide bond formation by an ATP-grasp enzyme follows the same acylphosphate logic regardless of the substrate pair: a carboxylate attacks the γ-phosphate of ATP to form an acylphosphate, and the incoming amine attacks that acylphosphate to give the amide. The review shows this reaction for a typical ATP-grasp enzyme. The practical limits are the ATP requirement and the narrow substrate scope; ATP-grasp enzymes are generally specialized for one physiological reaction rather than programmable like NRPSs. Their advantage is tractability. They are small relative to NRPS modules, and the conserved fold is well represented structurally, which makes them reasonable starting points for engineering, although the review reports no engineered variants.
The AET route is the most interesting of the five for cheap peptide production, and the least characterized. Kenzo and colleagues reported an enzymatic method using Empedobacter brevis ATCC 14234 that produces oligopeptides from unprotected amino acids in high yield. Using unprotected amino acids is significant because most chemical peptide synthesis spends its cost on protecting and deprotecting groups; an enzyme that accepts free amino acids removes that overhead. The catalyst identified in the strain was named carboxypeptidase Y, and it was reported to assist rapid oligopeptide production. However, the review notes that no amino acid sequence, coding gene sequence, or 3D crystal structure was provided for this enzyme, so the molecular identity of the catalyst remains unverified.
Isao ABE and colleagues later reported the first cloning and expression of α-amino acid ester acyltransferases from E. brevis and a second strain. The two AET amino acid sequences share 35% and 36% identity with α-amino acid ester hydrolase from Acetobacter pasteurianus, a level that places them in the same enzyme family but with substantial divergence. AETs display dual activities: they act as dipeptidyl peptidases, releasing dipeptides from the N-terminus of a substrate, and as transferases, transferring an aminoacyl or peptidyl group to a nucleophile. They are quite specific for both the acyl donor and the nucleophile, which is double-edged. Specificity gives predictable products, but it caps the substrate range.
The gap in this family is stark: no 3D structure and no reaction mechanism have been reported for AETs. That means rational engineering, structure-guided specificity tuning, and even reliable expression optimization have to proceed blind. For any laboratory considering this route, the practical implication is that the method works in the reported strain, but reproducing or improving it requires the original strain and empirical screening. Closing the structural gap is the obvious next step.
β-lactam antibiotics are defined by the four-membered β-lactam ring and include the penicillins, cephalosporins, and thiamycins. They are among the most widely used anti-infective agents, which makes their manufacturing chemistry a matter of scale. The review states that enzymatic synthesis of β-lactam antibiotics is increasingly applied in pharmaceutical manufacturing as an environmentally friendly and cost-effective route. That claim should be read carefully. It is an adoption argument from industrial practice, not a measured comparison. The review does not provide a cost or emissions benchmark against chemical synthesis. Still, the industrial record is real, and the enzyme class responsible is the β-lactam acylases.
These enzymes process β-lactam antibiotics and can be used to create semi-synthetic β-lactam antibiotics. The types studied are penicillin acylase PA , glutaryl acylase GA , and β-amino acid ester hydrolase AEH . Penicillin acylases are produced by many microorganisms and are categorized into two types based on substrate specificity. Their dominant industrial use is to produce 6-aminopenicillanic acid 6-APA , the active pharmaceutical intermediate from which semi-synthetic penicillins are made. The same enzymes can also catalyze acyl transfer to nucleophiles other than water, which is the activity exploited in peptide synthesis, racemic resolution, and production of chiral and achiral pharmaceutical intermediates.
From a user's standpoint, penicillin acylase is the most mature of the five routes discussed here. It is produced by many microorganisms, operates at industrial scale, and its dominant use, production of 6-APA, is the canonical example of enzymatic semi-synthetic antibiotic manufacturing. The review does not report a PA structure, but the enzyme is an amide hydrolase, and its synthetic value comes from transferring the acyl group to a nucleophile other than water. That transfer reaction is the basis of the peptide synthesis applications listed above, and it competes with hydrolysis of the acyl-enzyme intermediate by water, so reaction conditions determine whether the product is the peptide or the free acid.
Cyanophycin granule polypeptide CGP is an intracellular storage polymer found in most cyanobacteria. It contains equimolar arginine and aspartic acid, and the structure is unusual: arginine is linked through its α-amino group to the β-carboxyl group of each aspartic acid. The polymer is therefore a poly aspartic acid backbone with arginine side chains attached by isopeptide bonds, not a ribosome-made protein. In most cyanobacteria the cyanophycin synthetase gene cphA is identified and verified as responsible for CGP synthesis.
Degradation is handled by cyanophycinases. CphB catalyzes intracellular degradation of CGP, and CphE catalyzes extracellular degradation; both release the dipeptide β-Asp-Arg. Because the degrading enzymes trim the polymer to a single dipeptide product, the synthesis is simple in concept: produce CGP and its degrading enzyme simultaneously, then collect the dipeptide. The review presents this simultaneous CGP/CGPase production as an efficient route to β-Asp-Arg and illustrates the cyanophycinase reaction mechanism. Recombinant production of CGP has been established in several hosts:
| Host | Organism type | What was established |
|---|---|---|
| Escherichia coli | Bacterium | CGP production and isolation |
| Nicotiana tabacum | Plant | CGP production; later co-expression of CGP and CGPase |
| Pseudomonas putida | Bacterium | CGP production and isolation |
| Pseudomonas alcaligenes DIP1 | Bacterium | CGP production and isolation |
The plant work has moved furthest. A recent study achieved co-expression of CGP and CGPase in Nicotiana tabacum, and a follow-up suggested that tobacco can serve as a synthetic model for storage of arginine and for efficient transport of arginine and β-Asp-Arg dipeptides. The application target is feed or food that requires arginine content, where β-Asp-Arg could serve as the delivery form. Metabolic engineering of suitable hosts, together with chemo-enzymatic strategies, is presented in the review as a feasible route to dipeptides such as β-Asp-Arg. What remains open is scale. Plant-based production of a defined dipeptide at feed volumes, with extraction and purification, is not yet a demonstrated process.
The five families answer different questions, and the right choice depends on the product and the budget.
| Route | Activation, ATP cost | Best suited for | Key limitation | Structure reported |
|---|---|---|---|---|
| NRPS | Aminoacyl-AMP, ATP | Complex therapeutic peptides: penicillin, bleomycin, cyclosporine | Large modules, difficult engineering | Yes, PDB 2VSQ |
| ATP-grasp | Acylphosphate, ATP | Small peptides: D-Ala-D-Ala, glutathione, purine intermediates | Narrow substrate scope | Yes, PDB 2IP4 |
| AET | Ester acyl donor, no ATP | Short oligopeptides from unprotected amino acids | No sequence or structure reported | No |
| β-Lactam acylase | Amide hydrolysis or transfer, no ATP | 6-APA, semi-synthetic β-lactams, peptide synthesis | Transfer competes with hydrolysis | Not reported in review |
| Cyanophycinase | Isopeptide hydrolysis, no ATP | β-Asp-Arg from CGP | Requires upstream polymer production | Not reported in review |
For a target that is already a microbial natural product, NRPS-based whole-cell fermentation is the established path. For a small dipeptide made from cheap amino acids, the AET route is conceptually the most economical, because it avoids ATP and protecting groups, but it is also the least supported by evidence: the sequence and structure gaps make it a research project rather than a kit. For β-lactam intermediates, penicillin acylase is the industrially mature choice and the least risky. For β-Asp-Arg specifically, the cyanophycin route, especially plant-based co-expression, is the most promising but the least developed at scale.
Three unresolved questions define the field. First, what is the amino acid sequence, gene sequence, and 3D structure of the carboxypeptidase Y-like enzyme from E. brevis ATCC 14234? Without that, the highest-yield route to unprotected amino acid polymerization cannot be reproduced rationally. Second, what is the 3D structure and detailed reaction mechanism of the AETs? The 35% and 36% sequence identity to the Acetobacter hydrolase is the only molecular anchor available, and it is not enough for engineering. Third, how can CGP/CGPase co-expression be scaled for practical β-Asp-Arg production in feed or food? The tobacco model is elegant, but a process is not a demonstration. The review also leaves the details of NRPS amide bond formation largely at the schematic level, so the mechanistic floor for the most versatile family remains thin.
Peptides referenced: Glutathione.
Related reading: How Chameleon Cyclic Peptides Cross Membranes for Oral Drugs, Neoantigen Peptide Synthesis Services and GMP Manufacturing, Enzyme classes for biocatalytic synthesis of short oligopeptides, Choosing Coupling Reagents for Solid-Phase Peptide Synthesis.