Vancomycin Production: Processes, Challenges, and Quality Control

Vancomycin is produced by fermentation of the soil bacterium Amycolatopsis orientalis, followed by a labor-intensive purification process that must separate the active drug from a soup of closely related byproducts. Unlike small-molecule drugs that can be synthesized in a few chemical steps, vancomycin’s complex structure makes total chemical synthesis impractical at industrial scale. The result is a manufacturing chain where biology does the heavy lifting, and every stage from nutrient selection to final crystallization shapes whether the end product is pure enough for clinical use.

The Organism That Makes Vancomycin

Vancomycin comes from Amycolatopsis orientalis, a filamentous actinomycete found in soil. The bacterium was originally classified under Streptomyces and later Nocardia before landing in its current genus. Its genome runs about 9 million base pairs and encodes over 8,000 proteins, many of which are devoted to producing secondary metabolites the bacterium uses to compete with neighboring microbes.¹ Genome sequencing has identified 26 gene clusters related to secondary metabolism in a single production strain, but only one of those clusters, spanning roughly 64 kilobases and known as the vcm cluster, is responsible for vancomycin.²

That 64-kb stretch encodes everything the bacterium needs to assemble the vancomycin backbone, modify it, and add sugar residues. Experimental work using targeted mutagenesis has confirmed the roles of specific enzymes within the cluster, including a halogenase that attaches chlorine atoms, a methyltransferase, and two glycosyltransferases that add the sugar groups distinguishing vancomycin from its undecorated precursor.² The bacterium also carries many other secondary metabolite gene clusters, and as we will see later, silencing some of those competing clusters is one modern strategy for boosting vancomycin yield.

How the Bacterium Assembles Vancomycin

Vancomycin belongs to the glycopeptide antibiotic family, and its backbone is a heptapeptide, a chain of seven amino acid building blocks. What makes this chain unusual is that it is not built by ribosomes the way most proteins are. Instead, massive multi-domain enzymes called nonribosomal peptide synthetases (NRPSs) stitch the amino acids together in an assembly-line fashion. Studies on the closely related glycopeptide balhimycin showed that the NRPS machinery is encoded by four genes whose protein products collectively contain eight modules, even though the final product is only seven amino acids long.³ For the vancomycin-type antibiotic chloroeremomycin, researchers identified a three-subunit NRPS system and showed that even a single module expressed in E. coli could correctly recognize and activate its target amino acid, leucine.⁴

Once the linear heptapeptide chain is complete, it undergoes a series of oxidative cross-linking reactions that give vancomycin its rigid, cup-shaped three-dimensional structure. Three cytochrome P450 enzymes, designated OxyA, OxyB, and OxyC, catalyze these cross-links in a defined order. After the final biaryl bond is formed and the peptide is released from the enzyme, sugar groups are attached by glycosyltransferases GtfA and GtfB, and the N-terminal leucine residue is methylated by MtfA to complete the molecule. Each of these tailoring steps matters for biological activity: the sugars influence how tightly vancomycin binds its bacterial target, and the chlorine atoms added by the halogenase affect potency.

Fermentation Conditions and Media Design

Industrial vancomycin production depends on submerged fermentation, typically in stirred-tank bioreactors where A. orientalis grows as branching filamentous mycelium. Getting high yields is not simply a matter of feeding the organism and waiting. The carbon source, nitrogen source, mineral salts, phosphate concentration, and even the physical conditions of the bioreactor all shape how much vancomycin the culture produces versus how much energy it diverts into biomass growth or competing metabolites.

Carbon source selection illustrates the tradeoffs nicely. One study found that maltodextrin significantly increased vancomycin output, while glycerol promoted cell growth but did little for antibiotic synthesis.⁵ Yet a separate group working with a different strain reported that crude glycerol derived from biodiesel production was actually the most effective carbon source, achieving a vancomycin yield of about 7.6 g/L after 120 hours of flask fermentation, roughly 3.5 times higher than their baseline medium.⁶ The discrepancy is a reminder that strain-to-strain variation matters enormously: what works for one isolate of A. orientalis may not work for another.

Nitrogen sources show similar complexity. Soybean meal and soya flour have been identified as strong supporters of both biomass and vancomycin yield, while potato protein performed well in the glycerol-based system.⁵ Phosphate concentration is another lever: too little starves the culture, but too much can suppress secondary metabolite production in actinomycetes generally. This interplay between carbon, nitrogen, and phosphate means that media optimization usually relies on statistical experimental designs rather than one-variable-at-a-time testing.

Physical conditions inside the bioreactor add another layer of difficulty. Filamentous organisms like A. orientalis and their close relatives in Streptomyces are sensitive to agitation and aeration because these factors affect shear stress on the mycelium, oxygen transfer, and nutrient distribution. Too much shear can fragment the filaments and shift the culture toward growth rather than antibiotic production; too little can create oxygen-starved pockets where production stalls. Balancing these variables at pilot and production scale is one of the persistent engineering challenges of glycopeptide manufacturing.

Purifying Vancomycin From the Fermentation Broth

After fermentation, the broth is a messy mixture of cells, leftover nutrients, vancomycin, and dozens of related compounds the bacterium also produced. Getting pharmaceutical-grade vancomycin out of this mixture involves several steps, each with its own challenges.

The traditional route begins with filtration or centrifugation to remove cell mass, followed by adsorption of vancomycin from the clarified broth onto a resin. Polymeric adsorbents have been studied for this purpose, and researchers have explored whether vancomycin can be captured directly from unfiltered whole broth, which would skip the costly cell-removal step entirely.⁷ Direct-from-broth adsorption remains tricky in practice because cell debris and other macromolecules foul the resin, but it is an area of active process development.

Crystallization is often the final purification step and also the one where crystal form and purity are locked in. Vancomycin can crystallize in different habits depending on conditions. A study of salting-out crystallization using sodium chloride found that orthorhombic-type crystals dominated across a wide range of pH and salt concentrations, while needle-shaped crystals (generally less desirable for downstream processing) could be largely avoided by controlling pH and keeping salt concentration above a threshold where salting-in effects occur.⁸

Ion exchange resins have also been used to improve crystallization. By increasing the available surface area per unit volume inside the crystallization vessel, certain cation and anion exchange resins raised vancomycin yield above 97% while cutting the crystallization time in half, from about 24 hours to 12 hours. Purity stayed at roughly 95% regardless of the added surface area, suggesting that the resin accelerated crystal formation without introducing new impurities.⁹

The Impurity Problem

Because vancomycin is made by a living organism performing dozens of enzymatic reactions, minor variations in those reactions create structural cousins of the drug that end up in the final product. Some of these are well-characterized: norvancomycin (missing the N-methyl group), desvancosaminyl vancomycin (missing a sugar), and the dechlorinated variants (missing one or both chlorine atoms). Others are degradation products that form during processing or storage. Separating and quantifying these impurities is one of the more analytically demanding tasks in antibiotic quality control.

High-performance liquid chromatography (HPLC) is the workhorse method. Early HPLC methods struggled to resolve vancomycin from its closest relatives, particularly the monodechlorovancomycin variants, which differ from the parent molecule by only one chlorine atom. A method development effort testing more than 40 reversed-phase columns found that most could not separate didechlorovancomycin from one of the monodechloro variants. Ultimately, a separation at pH 7.7 using dioxane as an organic modifier on a specific column type succeeded in resolving components that had never been fully separated before.¹⁰

Hydrophilic interaction chromatography (HILIC) offers a complementary approach. Because vancomycin and its relatives are polar molecules with sugar groups, they often behave poorly on standard reversed-phase columns. HILIC, which retains polar compounds more strongly, has been verified to provide good retention and selectivity for glycopeptide impurities including norvancomycin, desvancosaminyl vancomycin, didechloro vancomycin, and the crystalline degradation product known as CDP-1.¹¹

When entirely new impurities show up, identifying them requires heavier analytical firepower. In one case, an unknown impurity appearing at about 0.5% of the total was isolated using two-dimensional preparative liquid chromatography and then characterized by mass spectrometry and nuclear magnetic resonance. It turned out to be a vancomycin analog in which one amino acid residue on the side chain, N-methyl-leucine, had been swapped for N-methylmethionine, presumably because the NRPS machinery occasionally grabs the wrong amino acid.¹² Finding and identifying impurities at the sub-percent level matters because even trace contaminants in an antibiotic administered intravenously can contribute to adverse reactions or interfere with potency.

How Clean Are Commercial Products

A systematic quality assessment of parenteral vancomycin products sold in the United States found that the active component, vancomycin B, made up about 90 to 95% of each product. Total impurities ranged from roughly 5 to 11% depending on the analytical method used, with no single impurity exceeding 2%, and the degradation product CDP-1 at or below 2% across all tested products. All products surpassed United States Pharmacopeia acceptance criteria. That might sound reassuring, but the presence of 5 to 10% impurity material in an injectable drug is high compared with many small-molecule antibiotics, and it underscores why analytical methods capable of resolving individual impurities are so important.

CDP-1 deserves special attention. This crystalline degradation product forms when vancomycin sits at body temperature for extended periods, a situation that arises in patients with severe kidney impairment because the drug’s half-life is dramatically prolonged. The problem is compounded by the fact that common clinical assays used to measure vancomycin levels in patients’ blood, such as fluorescence polarization immunoassay, cross-react with CDP-1. In patients with end-stage kidney disease, this cross-reactivity has been shown to overestimate true vancomycin concentrations by 40 to 53%, potentially leading clinicians to underdose patients who actually need more drug. Newer mass-spectrometry-based assays and improved HPLC methods developed for production quality control have also found application in clinical laboratories trying to avoid this pitfall.¹³

Genome Editing to Boost Yields

Traditional strain improvement for antibiotic-producing organisms relied on random mutagenesis, exposing cells to UV light or chemical mutagens and screening thousands of colonies for higher producers. That approach works but is slow and unpredictable. CRISPR-based genome editing has opened a more targeted route.

In one study, researchers applied a CRISPR/Cas9 system directly in a vancomycin-producing strain of Amycolatopsis to delete gene clusters responsible for a competing secondary metabolite called ECO-0501. The logic was straightforward: if the bacterium is spending resources making a compound you do not want, removing those genes should free up precursors and cellular machinery for vancomycin. It worked. Mutants with portions of the ECO-0501 gene cluster deleted showed a 30 to 40% increase in vancomycin yield.¹⁴

Separate work using a CRISPR-Cas12a system in a related Amycolatopsis species demonstrated that markerless, scarless gene deletions are now feasible in these organisms, a technical milestone since Amycolatopsis has historically been difficult to genetically manipulate.¹⁵ That study targeted vanillin biosynthesis rather than vancomycin, but the genome-editing toolkit it developed applies to any secondary metabolite pathway in the genus. The ability to make precise deletions without leaving antibiotic-resistance markers behind is especially important for strains destined for pharmaceutical manufacturing, where regulatory agencies scrutinize every foreign DNA element in the production organism.

These CRISPR approaches could be combined: silencing multiple competing metabolite clusters in the same strain, fine-tuning regulatory genes that control when and how much vancomycin the bacterium produces, or even swapping in modified NRPS modules to produce vancomycin analogs with improved properties. The field is still early, but the trajectory points toward engineered strains that produce more vancomycin, with fewer byproduct impurities, in shorter fermentation runs.

Environmental Concerns From Manufacturing Wastewater

Pharmaceutical manufacturing wastewater containing antibiotic residues is a growing environmental concern, and vancomycin production is no exception. A study of two pharmaceutical wastewater treatment plants in eastern China found that their treatment systems removed over 99% of vancomycin from the influent, primarily through biodegradation in aerobic tanks rather than sorption onto sludge.¹⁶ That sounds like good news, but the concentrations in the effluent were still high enough that environmental risk quotients exceeded the safety threshold of 1, meaning the discharged water posed non-negligible risks to aquatic organisms and potentially to the spread of antibiotic resistance.

Antibiotic residues in the environment at sub-lethal concentrations can exert selective pressure on environmental bacteria, promoting the development and spread of resistance genes. For vancomycin, this is an especially fraught issue because vancomycin-resistant enterococci (VRE) are already a serious clinical problem, and any environmental reservoir of resistance genes could accelerate their spread. Manufacturing facilities that produce vancomycin at scale face regulatory pressure to invest in advanced treatment technologies, such as advanced oxidation or activated carbon polishing, to bring effluent concentrations below ecologically relevant thresholds. The treatment challenge is compounded by the fact that vancomycin’s large molecular size and hydrophilic character make it resistant to some conventional removal methods, even as biodegradation in well-operated aerobic systems proves effective.

Why Vancomycin Remains Difficult to Replace With Synthetic Chemistry

Total synthesis of vancomycin has been achieved in academic laboratories multiple times since the late 1990s, but every published route requires dozens of synthetic steps and produces the drug in quantities measured in milligrams, not kilograms. The molecule’s complexity is the obstacle: it has multiple stereocenters, unusual amino acids, cross-linked aromatic rings, two sugar groups, and a chlorine atom on each of two aromatic rings. Reproducing all of that chemistry in a flask, in the right sequence, with acceptable yields at each step, makes fermentation look efficient by comparison.

Semi-synthetic approaches, where the fermentation product is chemically modified to create new analogs, have been more practical. Drugs like oritavancin and telavancin are semi-synthetic glycopeptides derived from vancomycin-type scaffolds. But even these depend on a fermentation-derived starting material. For the foreseeable future, producing vancomycin itself at pharmaceutical scale means growing A. orientalis in a bioreactor and then cleaning up what it makes. The focus of innovation, as the CRISPR work and media-optimization studies show, is on making that biological process faster, cheaper, and cleaner rather than trying to bypass it altogether.