Phages, lysins, antimicrobial peptides (AMPs), and engineered antibacterial biologics differ markedly in how they work, but all face the same practical requirement: they must become viable pharmaceutical products. Promising activity is only the beginning. A candidate must also be produced reproducibly, purified sufficiently, characterized meaningfully, formulated appropriately, stored without unacceptable loss of activity, and tested against defensible quality standards.
For phage therapies, that chain of requirements is becoming increasingly explicit. A European Medicines Agency (EMA) draft guideline addresses the manufacture, control of materials, characterization, specifications, analytical control, reference standards, and stability of bacteriophage active substances, as well as the development, manufacture, control, and stability of finished products.1 Its scope shows how much pharmaceutical work separates an active phage from a finished therapeutic product.
Similar questions apply to other biological antimicrobials, even when their production processes differ.
Manufacturability therefore cannot remain a concern addressed only after a promising antimicrobial has emerged from discovery. Production behavior, purification, stability, formulation, and analytical tractability can all influence whether biological activity can ultimately be translated into a repeatable pharmaceutical product.
Manufacturability Begins with Candidate Selection
Lead selection traditionally emphasizes whether a candidate produces the desired biological effect. For new antimicrobial modalities, it can also account for whether a molecule or biological agent has properties compatible with the process required to manufacture it.
Phages make the relationship between biology and manufacturing particularly clear. Their manufacture begins with the creation and maintenance of bacterial and phage banks, followed by production of the phage in its bacterial host, downstream purification to produce drug substance, and subsequent formulation and filling. The upstream process therefore depends directly on the phage-host system used for propagation.2
Engineering can expand the number of possible candidates even further, particularly for modular biologics whose sequences or domains can be deliberately recombined or modified. One high-throughput lysin platform used an iterative design-build-test approach to generate approximately 10,000 engineered variants and identify candidates active against multidrug-resistant Acinetobacter baumannii. Initial screening incorporated both antibacterial performance and expression yield, and results from early libraries were used to enrich subsequent rounds before a lead was selected that combined strong expression with activity in human serum and an ex vivo burn-wound model. The study demonstrates how molecular engineering can create a large candidate space in which properties relevant to both biological performance and eventual production can influence which molecules advance.3
Once multiple variants demonstrate suitable antibacterial activity, attributes such as expression, soluble recovery, stability, purification compatibility, and formulation can become part of the selection logic rather than problems addressed only after the molecular design has been fixed. Incorporating those considerations earlier can reduce the risk of advancing candidates whose technical liabilities emerge only after substantial work has occurred.
Phages Expose the Full Complexity of Biological Manufacturing
Bacteriophages offer perhaps the clearest example of how the biology of an antimicrobial can dictate its manufacturing process. A typical sequence includes development of a master cell bank and master phage bank, propagation of the phage in its bacterial host, and downstream purification to produce drug substance before formulation into the final drug product.2
Phage propagation depends on infection and lysis of the production host. That mechanism creates a significant downstream burden because bacterial rupture releases proteins, lipids, nucleic acids, and other host-derived materials into the crude lysate. Producing purified drug substance can consequently require multiple stages of clarification, capture, concentration, purification, and polishing rather than a single separation step.
Individual phages may also respond differently to purification operations. Centrifugation, filtration, chromatography, and other processing approaches can produce different recoveries or losses depending on the phage, limiting the extent to which one purification process can simply be transferred unchanged from candidate to candidate.
Phage cocktails add another level of complexity. A therapeutic combination may require the production of multiple constituent phages, while the stability of a cocktail can vary with its composition. Manufacturing strategies therefore must accommodate several biological components rather than optimizing exclusively around a single active agent.4
A platform can standardize the development framework without eliminating phage-specific process work. Common banking strategies, upstream workflows, purification-development tools, and quality practices can provide a reusable structure, while individual process steps may still need to be adapted to a particular phage or phage combination.
Engineered phages introduce additional control considerations. Development must establish the identity and purity of phage banks and determine whether mutations have occurred, adding another dimension to the characterization of the material entering manufacturing.
Lysins Turn the Challenge into Recombinant Protein Development
Lysins remove some of the complexities associated with propagating an intact bacteriophage, but they introduce a different set of manufacturing constraints. Because these antibacterial enzymes can be produced recombinantly, their manufacture can draw on recombinant protein-expression and purification approaches. That does not make production straightforward.
Soluble expression is one potential limitation. In a study examining two recombinant endolysins expressed in Escherichia coli, only about 30–40% of total endolysin was ordinarily recovered in the soluble fraction. Changes to induction conditions, fusion tags, chaperone co-expression, osmolytes, and extraction conditions were examined in an effort to increase recovery.5
The results illustrate why optimization can require empirical testing. The addition of N-lauroylsarcosine during extraction increased recovery to 54% of total endolysin content, demonstrating that expression and extraction conditions can materially affect how much usable protein is obtained.
Those expression challenges do not mean lysins are inherently confined to small-scale production. In one example, a glucose-limited fed-batch process for the lysin TSPphg was demonstrated at 20 L, producing 3,322.8 ± 26 mg/L bioactive protein at 95.5 ± 0.7% purity after 26 hours of fermentation. The result provides a concrete example of how an optimized recombinant process can move a lysin beyond small laboratory cultures, while underscoring that successful scale-up depends on solving candidate-specific expression and purification constraints first.6
Another recombinant lysin example illustrates how expression, purification, and formulation can be linked across a development program. LysB4 was recovered using one-step cation-exchange chromatography at greater than 90% purity, with approximately 132 mg obtained from one liter of culture. Formulation screening then evaluated conditions intended to maintain stability and reduce aggregation.7
These examples shift the question away from whether lysins can be manufactured at all to whether a particular molecule can be expressed, recovered, purified, and stabilized efficiently enough to support the intended product.
That matters especially when engineering can generate large numbers of potential variants. Once antibacterial performance has identified multiple viable candidates, production behavior becomes another characteristic that can help determine which molecules are practical to advance.
Antimicrobial Peptides Force an Early Production-Route Decision
AMPs pose another version of the manufacturability problem because developers may have multiple routes for producing the same or related molecules. Chemical synthesis provides precise structural control but can become expensive and complex at scale, while heterologous expression offers a potentially scalable alternative that requires its own process optimization.8
Two recurring obstacles in recombinant AMP production are limited yield and proteolytic degradation. The biology of the molecule can create another problem: an antimicrobial peptide expressed inside a microbial production host may be harmful to the organism being used to manufacture it.8,9
Fusion strategies provide one way to address that tension. In one recombinant production system, small ubiquitin-like modifier (SUMO) fusion enabled production of intact host-defense peptides while limiting toxicity to the bacterial expression host. The process incorporated a simplified two-step purification scheme and was demonstrated in a 10-L pilot-scale fermentation.9
These examples show why production route should be considered alongside molecular design. A sequence intended for chemical synthesis presents a different set of process constraints from one intended for recombinant expression, while a biological route must contend with expression level, degradation, host compatibility, and purification.
The synthesis-versus-expression decision therefore belongs within candidate and process design rather than after the molecule has effectively been fixed. The optimal production strategy may influence which version of an AMP is most practical to advance.
Formulation Is Part of Molecule Development
Producing purified material is only part of the path to a usable antimicrobial product. The drug substance must also survive formulation, storage, handling, and administration while retaining the characteristics required for treatment.
Phage development illustrates the risks of addressing this too late. Formulation has often been considered only after substantial drug-substance and manufacturing work has already occurred, an approach that can lengthen timelines and make required product-performance parameters more difficult to achieve.10
Formulation serves two closely connected purposes for phages: delivering active particles in an appropriate dosage form and maintaining stability over storage. Formulation choices also vary with the intended route and presentation, including liquid, solid, and semisolid products.
Individual phages may not behave identically under the same conditions, and the stability of phage cocktails can also vary. A formulation strategy therefore must preserve acceptable performance for the individual phage or combination being developed rather than assume universal compatibility.
Lysins demonstrate the same principle in a recombinant protein context. Formulation screening for LysB4 identified conditions that maintained nearly complete integrity and biological activity for 12 weeks at 4 °C and −20 °C, while significant degradation occurred after three weeks at 25 °C.7
AMPs present a different set of formulation problems. Stability, cytotoxicity, and bioavailability have been identified as barriers to clinical translation, and investigated delivery approaches include micelles, liposomes, hydrogels, inorganic nanoparticles, chemical conjugation, targeted delivery, and responsive-release systems.11
Across these modalities, formulation is not simply a means of placing an already finished molecule into a dosage form. It can determine whether activity can be maintained through storage and delivery. Beginning formulation work while there is still flexibility to modify the candidate, process, or presentation can expose limitations before major process and product decisions become difficult to change.
New Modalities Need Analytics That Reflect Their Biology
Manufacturing a complex biological antimicrobial consistently requires more than confirming that material has been produced. Developers need analytical approaches capable of establishing product identity and quality and determining whether relevant biological activity has been maintained.
For phages, that problem extends across process development and quality control. Analytical testing is required to determine whether drug substance and drug product meet defined specifications, while established approaches for quantifying infective phages can be laborious and time-consuming.
The developing regulatory framework makes the analytical burden explicit. The EMA guideline addresses characterization, control of the active substance, analytical considerations, reference standards or materials, finished-product control, and stability, embedding analytical work throughout the quality strategy rather than placing it solely at final release.
The general principle extends beyond phages. As nontraditional antimicrobial biologics move toward clinical and commercial development, they still require a defined quality framework for determining whether drug substance and drug product are acceptable. The International Council for Harmonisation (ICH) Q6B guideline provides general principles for setting and justifying specifications for biotechnology and biological products, with specifications comprising the tests, analytical procedures, and acceptance criteria used to evaluate product quality.12 For emerging antimicrobial modalities, that means analytical development ultimately has to translate unusual biological properties into measurable attributes that can be controlled consistently from batch to batch.
Phages also illustrate why the analytical strategy must reflect the biology of the product. The presence of physical phage material does not by itself establish infective activity. Manufacturing and analytical work therefore have to remain closely connected as processing, formulation, and storage conditions are established.10
For emerging modalities, the broader implication is that analytical methods should measure attributes that matter to product identity, quality, and biological performance rather than rely narrowly on the quantity of material present.
Engineer for Developability, Not Just Potency
The ability to engineer antimicrobial biologics creates an opportunity to address manufacturability earlier rather than accept it as a fixed property of a discovered molecule.
Engineered lysins provide a useful model. High-throughput platforms can generate thousands of variants, creating a broad design space before a final lead is selected. At the same time, recombinant lysin studies demonstrate that soluble expression and recovery can become meaningful constraints.
AMP production presents a related opportunity. Fusion technologies can mitigate some of the difficulties created by recombinant expression, including toxicity to the host and subsequent purification requirements, while broader AMP-production research evaluates approaches intended to improve yield and stability alongside biological performance.
Developability can therefore become a formal selection criterion rather than a problem evaluated only after lead nomination. Once several candidates demonstrate adequate antibacterial activity, differences in how readily they can be produced, recovered, stabilized, and formulated can help distinguish the molecules that are more practical to advance.
This changes the relationship between discovery and manufacturing. Rather than fixing the candidate first and asking process teams to accommodate whatever liabilities emerge, production and formulation data can help determine which candidate becomes the lead. For modalities already being engineered, that same design loop can incorporate manufacturability alongside antibacterial performance.
Building Platforms Without Forcing Uniformity
New antimicrobial modalities will benefit from manufacturing platforms, but useful platforms will need to accommodate biological variability rather than assume it away.
Phage production illustrates the distinction. A common framework can include controlled cell and phage banks, defined upstream workflows, purification-development approaches, formulation work, analytical testing, and quality control. Individual phages can nevertheless respond differently to processing and purification, and cocktails create additional product-specific requirements.
A useful platform therefore does not have to mean a single fixed process. Standardization can instead focus on how decisions are made: how candidate production is assessed, how purification conditions are screened, how formulation work is initiated, and how appropriate analytical controls are established. Those workflows can be reusable even when individual molecules require different operating conditions.
The emerging regulatory framework for phage therapy reinforces the value of this continuity. By addressing drug substance manufacture, material controls, characterization, analytical testing, pharmaceutical development, finished-product manufacture, and stability within one quality framework, it connects the properties of the biological starting material to the requirements of the final medicinal product.
Novel antibacterial activity can identify a promising candidate, but a viable medicine must also become a reproducible product. The goal is not to force every phage, lysin, or peptide into the same manufacturing process, but to make production, characterization, formulation, stability, and quality part of deciding which candidates advance.
