Five enzyme families make short oligopeptides and peptide antibiotics without ribosomes: nonribosomal peptide synthetases, ATP-grasp ligases, alpha-amino acid ester acyltransferases, beta-lactam acylases, and cyanophycinases. Each activates its substrate differently, and the evidence behind each…
The answer to the reader's question is that enzymatic synthesis of oligopeptides and peptide antibiotics runs through five enzyme families, each solving the same chemical problem in a different way: making a carboxyl carbon reactive enough to accept an amine nucleophile and form an amide bond. Nonribosomal peptide synthetases NRPS assemble peptides on modular protein templates. ATP-grasp enzymes , also called ATP-dependent carboxylate-amine ligases, activate carboxylic acids as acylphosphate intermediates. α-amino acid ester acyltransferases AETs move an acyl group from an ester donor onto an amino nucleophile. β-lactam acylases cleave and rebuild the amide bonds of β-lactam antibiotics at industrial scale. Cyanophycinases degrade a cyanobacterial storage polymer to release the dipeptide β-Asp-Arg. The five-way division and every mechanism described here trace to a 2019 review by Wang et al. in Biomolecules, volume 9, issue 11, article 733.
That review divides the biocatalysts that make nonribosomal peptides into two groups by how they activate their substrates. ATP-independent enzymes, including transacylases, use aminoacyl phosphate . ATP-dependent enzymes, including tRNA-dependent ligases, use aminoacyl-adenosine monophosphate . ATP-grasp enzymes occupy a mechanistically distinct position: they consume ATP to form an acylphosphate intermediate and, in most cases, require an Mg2+ ion coordinated by ATP in the active site.
| Enzyme family | Substrate activation | Structural signature | Reported example | Structure or figure | Stated limitation |
|---|---|---|---|---|---|
| Nonribosomal peptide synthetases NRPS | ATP-dependent aminoacyl-AMP or ATP-independent aminoacyl phosphate | A, C, Te, and PCP domains | Penicillin, bleomycin, cyclosporine | PDB 2VSQ, Fig. 1 and Fig. 2 | No step-by-step chemistry; the two activation routes are not compared |
| ATP-grasp enzymes | ATP-dependent acylphosphate intermediate | Three conserved domains; nonclassical ATP-binding fold; Mg2+ coordinated by ATP | Glycinamide ribonucleotide synthetase | PDB 2IP4, Fig. 3 | Role of Mg2+ beyond ATP coordination is unresolved |
| α-amino acid ester acyltransferases AETs | Not established | None solved | Oligopeptides from unprotected amino acids using Empedobacter brevis ATCC 14234 | Fig. 4 | No 3D structure or mechanism; yields are qualitative |
| β-lactam acylases | Acyl transfer at the β-lactam amide | PA, GA, and AEH classes; two PA types by substrate specificity | 6-APA, semi-synthetic β-lactams | No figure in the source | The two PA types are unnamed |
| Cyanophycinases CphB, CphE | Hydrolytic degradation of CGP | Intracellular and extracellular cyanophycinases | β-Asp-Arg dipeptide | Fig. 5 | No quantitative CGP or β-Asp-Arg yields |
One caution applies to everything below. The source is a review, not primary experimental work, and several catalytic mechanisms appear only as figures without step-by-step chemistry. Where numbers are missing, this article says so rather than supplying them.
Nonribosomal peptide synthesis is a widespread and essential biochemical process in bacteria and fungi, and it produces some of the most valuable peptide therapeutics in use. Penicillin, bleomycin, and cyclosporine are all NRPS products, and together they show how far the family's structural and biological range extends.
A typical NRPS enzyme contains four domains: A adenylation , C condensation , Te thioesterase , and PCP peptidyl carrier protein . The structure of a representative enzyme is deposited as PDB ID 2VSQ and shown in Fig. 1. NRPS enzymes catalyze amide bond formation Fig. 2 . The review does not walk through the condensation chemistry step by step, so the domain architecture is far better documented than the catalytic detail.
The activation chemistry splits cleanly. Transacylase-type enzymes work without ATP and activate substrates as aminoacyl phosphate. tRNA-dependent ligase-type enzymes consume ATP and activate substrates as aminoacyl-adenosine monophosphate. Which route is more efficient, and which covers a broader product range, is not settled.
The clinical footprint of NRPS products is real even where the enzymology is not. Bleomycin, used in cancer chemotherapy, was reported to cause pruritic, erythematous linear flagellate dermatitis on the lower back of a man with stage 3 seminoma after his third cycle of bleomycin, etoposide, and cisplatin; his pruritus and erythema improved with bilastine and desoximetasone PMID 34484791 . That case report documents a toxicity of an NRPS product. It says nothing about how the molecule is made.
ATP-grasp enzymes, also called ATP-dependent carboxylate-amine ligases, activate carboxylic acids as acylphosphate intermediates. They appear in diverse biological systems, including de novo purine biosynthesis, which makes them both a mechanistic model and a source of biosynthetic steps that intersect with nucleotide metabolism.
The family's better-known members include biotin carboxylase , Ddl D-alanine:D-alanine ligase , and glutathione synthetase . Structurally, ATP-grasp enzymes typically carry three conserved domains and a nonclassical ATP-binding fold that encloses the ATP molecule. Most require an Mg2+ ion coordinated by ATP in the active site.
Glycinamide ribonucleotide synthetase is the representative enzyme used to illustrate the family, and its structure is deposited as PDB ID 2IP4 Fig. 3 . ATP-grasp enzymes catalyze peptide bond formation, which places them alongside NRPS enzymes as enzymatic routes to amide bonds. What Mg2+ contributes beyond holding ATP in place remains an open mechanistic question, and it matters for anyone trying to redesign an active site.
Isao ABE and colleagues first reported cloning and expression of α-amino acid ester acyltransferases AETs from Empedobacter brevis and a second strain. The two AET amino acid sequences were 35% and 36% identical to an α-amino acid ester hydrolase from Acetobacter pasteurianus, which places them in a recognizable hydrolase neighborhood without identifying their catalytic residues.
AETs carry dual dipeptidyl peptidase and transferase activities, and they are highly specific for both their acyl donors and their nucleophiles. The reactions they catalyze are collected in Fig. 4. No study has yet examined the three-dimensional structure of an AET or its reaction mechanism, so the specificity is measurable while its structural basis is not.
A separate line of work, attributed to Kenzo and colleagues, reported an efficient enzymatic method for producing oligopeptides from unprotected amino acids with high yield using Empedobacter brevis ATCC 14234 . An enzyme catalyst named carboxypeptidase Y in that strain assisted rapid formation of certain oligopeptides. The amino acid sequence, the coding gene sequence, and the three-dimensional crystal structure of that carboxypeptidase Y were not provided.
The appeal of this route is easy to state and hard to quantify. Working from unprotected amino acids removes the protection and deprotection steps that dominate chemical oligopeptide synthesis, and the reported yields are described as high. But the yield claim is qualitative. No numerical values accompany it, no structure explains it, and the review presents no independent replication.
β-lactam antibiotics all contain a β-lactam ring , and the group includes penicillins, cephalosporins, and thiamycins. They are among the most widely used anti-infective agents and carry major importance in the pharmaceutical industry.
Enzymatic synthesis of β-lactam antibiotics is increasingly adopted in modern pharmaceutical companies because it is considered more environmentally benign and less costly than the alternatives. β-lactam acylases have historically been used to process β-lactam antibiotics, and they can also help produce semi-synthetic β-lactams. Three types have been studied for β-lactam biosynthesis: penicillin acylase PA , glutaryl acylase GA , and β-amino acid ester hydrolase AEH .
Penicillin acylases are produced by various microorganisms and are divided into two types based on substrate specificity; the review does not name those two types, which limits how precisely the classification can be applied to a new substrate. Industrially, penicillin acylases drive production of 6-aminopenicillanic acid 6-APA and semi-synthetic antibiotics, and they may support development of novel drugs. Beyond antibiotics, penicillin acylases can be used in peptide synthesis, in resolution of racemic mixtures, and in production of achiral and chiral compounds for pharmaceutical intermediates.
The historical record supports the 6-APA role. A review of early work in the penicillin series reported that the search for more potent derivatives culminated in the discovery of 6-aminopenicillanic acid as a penicillin biosynthesis intermediate PMID 1652802 . Process modeling has followed the enzyme into immobilized reactors: a theoretical study obtained semi-analytical steady-state solutions for the nonlinear reaction-diffusion equations describing immobilized penicillin G acylase using the Akbari-Ganji and modified Adomian decomposition methods, and those solutions agreed well with numerical results across the parameter range PMID 38045190 . That is a modeling result, not a manufacturing yield, and it should be read that way.
Clinical evidence around β-lactams measures use and allergy, not enzymatic synthesis. An outpatient nurse-driven penicillin allergy testing program in oncology screened 82 patients with reported penicillin allergy, found 90% eligible, and found 97% of those tested negative; aztreonam use among admitted hematopoietic stem cell transplant patients decreased compared with before the program PMID 33739344 . A trial has tested tailored Helicobacter pylori eradication based on clarithromycin resistance NCT03884348 , and a large neonatal trial is evaluating chlorination to reduce enteric and antibiotic-resistant infections NCT06824350 . Prescribing data show how much depends on clinician behavior: a prospective cohort of 145 Norwegian general practitioners found antibiotics prescribed in 27% of respiratory tract infection consultations, with penicillin V at 37% of prescriptions and macrolides at 28% PMID 19929185 . None of these studies evaluates an acylase, an AET, or an ATP-grasp enzyme, and none should be cited as evidence about enzymatic manufacturing.
Cyanophycin granule polypeptide CGP , also called multi-L-arginyl-poly, is an intracellular storage polymer found in most cyanobacteria. It contains equimolar arginine and aspartic acid, and the arginine moieties are linked through their α-amino group to the β-carboxyl group of each aspartic acid in the aspartic acid backbone.
The cphA gene, which encodes cyanophycin synthetase , has been identified and verified for CGP synthesis in most cyanobacteria genera. Degradation runs through two cyanophycinases , CphB and CphE , which act intracellularly and extracellularly to release β-Asp-Arg dipeptides. The reaction mechanism of the cyanophycinases is shown in Fig. 5.
Generating CGP and a cyanophycinase at the same time can efficiently yield β-Asp-Arg, which may be useful where feed or food needs added arginine. CGP production and efficient isolation have been established in recombinant hosts spanning bacteria and plants.
| Host | Organism type | Reported status for CGP |
|---|---|---|
| Escherichia coli | Bacterium | CGP production and efficient isolation established |
| Nicotiana tabacum | Plant | CGP and cyanophycinase co-expression achieved; arginine and β-Asp-Arg storage and transport reported |
| Pseudomonas putida | Bacterium | CGP production and efficient isolation established |
| Pseudomonas alcaligenes DIP1 | Bacterium | CGP production and efficient isolation established |
Co-expression of CGP and its degrading enzyme in Nicotiana tabacum was recently achieved, and a further study indicated that sufficient storage and efficient transport of arginine and β-Asp-Arg dipeptides is possible in this synthetic model. The conclusion that dipeptides such as β-Asp-Arg can be produced through metabolic engineering of suitable hosts and chemo-enzymatic strategies is plausible and supported in principle. It is not supported by quantitative yield data, because the review supplies none for CGP production or for β-Asp-Arg.
The five families are not interchangeable, and the evidence supporting each one differs in kind.
| Goal | Best supported route | Evidence status |
|---|---|---|
| Industrial 6-APA and semi-synthetic β-lactams | Penicillin acylases | Established industrial use per the review |
| Short oligopeptides from unprotected amino acids | AET and carboxypeptidase Y in Empedobacter brevis ATCC 14234 | Reported high yield, no numerical values, no structure |
| β-Asp-Arg dipeptide for arginine enrichment | Co-expressed CGP and cyanophycinase | Demonstrated in plant and bacterial hosts, no quantitative yields |
| Defined peptide bond formation for study or engineering | ATP-grasp ligases and NRPS | Solved or representative structures, defined activation chemistry |
| New β-lactam scaffolds | β-lactam acylases PA, GA, AEH | Possible per the review, classification incomplete |
For a research group, the practical move is to separate mechanistically characterized enzymes from reportedly productive ones. ATP-grasp ligases and NRPS enzymes have solved or representative structures PDB 2IP4 and 2VSQ and defined activation chemistries, which makes them suitable for engineering and for teaching the chemistry. AETs have neither a structure nor a mechanism, so work with them starts from activity assays and de novo structure determination. Penicillin acylases sit at the opposite end: decades of industrial use, an unnamed two-type classification, and a modeling literature on immobilized reactors.
For a buyer, three questions separate a real enzymatic route from a claim. First, what is the enzyme, with sequence and structure if available. Second, what is the quantitative yield, titer, and enantiomeric purity, since "high yield" without a number cannot be compared across routes. Third, at what scale were the data generated, because a plant co-expression model that stores and transports a dipeptide is not a commercial fermentation process. The review provides no cost data, so the claim that enzymatic β-lactam synthesis is cheaper should be attributed to the review's summary rather than treated as measured economics.
The gaps are specific and worth stating plainly, because they define where new work would be most useful.
Two further limits apply to the surrounding literature. Registry searches for peptide-adjacent questions return unrelated records: one registered study of renal fatty acid uptake in idiopathic uric acid nephrolithiasis was withdrawn with an enrollment of zero NCT02975986 , which is a reminder to check status and enrollment before citing a trial. And the clinical studies cited here measure antibiotic prescribing, allergy testing, and infection prevention, not enzyme performance. The enzymatic synthesis case rests on the 2019 review, on structural deposits, and on industrial practice. It does not yet rest on quantitative, replicated, head-to-head comparisons of the five families.
Peptides referenced: Glutathione.
Related reading: Condensation Agents in SPPS: How to Choose the Right One, Peptide Antigen Design: Key Parameters and Practical Guidelines, Why Custom Peptide Assays Fail: Six Preventable Causes, Enzymatic Synthesis of Oligopeptides: Five Enzyme Families.