Inside BioNTech’s Little-Known Phage-Derived Antibacterial Program
BioNTech is known globally for mRNA vaccines and, increasingly, for an expanding oncology pipeline. Much less visible is another technological programme inside the German biotechnology company: the development of precision antibacterials derived from bacteriophage biology.
The programme originated with BioNTech's acquisition of Austrian biotechnology company PhagoMed Biopharma in October 2021. BioNTech paid an upfront consideration of €50 million, less acquired debt, while former PhagoMed shareholders were eligible for up to another €100 million if defined clinical-development milestones were achieved. PhagoMed subsequently became BioNTech R&D Austria, providing the company with an antibacterial research operation in Vienna and a platform for engineering bacteriophage-derived lysins.
Rather than administering intact bacteriophages capable of replicating inside bacteria, this technology uses one of the molecular weapons phages normally deploy at the end of their infection cycle. Endolysins are enzymes produced by bacteriophages to break bacterial peptidoglycan, allowing newly assembled viral particles to escape from the infected cell. When appropriately engineered and delivered from outside the bacterium, these enzymes can themselves become antibacterial agents.
BioNTech calls its platform LysinBuilder. The company currently describes it as a system for engineering synthetic lysins capable of selectively targeting individual bacterial groups, with the broader ambition of creating precision antibacterials for persistent or drug-resistant infections while avoiding some of the collateral damage that broad-spectrum antibiotics can impose on the microbiome. Despite changes in BioNTech's clinical priorities during 2026, the LysinBuilder platform and synthetic lysin approach remain described on the company's current technology and infectious-disease pages.
The most advanced product to emerge from this work was BNT331, a recombinant endolysin developed against bacterial vaginosis. The candidate originated from PhagoMed's PM-477 programme and was designed primarily to eliminate Gardnerella, a group of bacteria strongly associated with bacterial vaginosis and with structured biofilms on vaginal epithelial surfaces.
The biological rationale differs fundamentally from conventional antibiotic treatment. Bacterial vaginosis is associated with disruption of a lactobacillus-dominated vaginal ecosystem and expansion of bacterial communities that frequently include Gardnerella. Standard antibiotics can reduce bacterial burden, but recurrence is common and treatment can also affect organisms that researchers would prefer to preserve. A sufficiently selective lysin could theoretically attack important BV-associated bacteria while leaving much of the surrounding microbiome intact.
Preclinical work established the principle. Earlier experiments with PM-477 showed rapid activity against diverse Gardnerella isolates and disruption of Gardnerella-dominated biofilms. BioNTech later evaluated an optimized BNT331 endolysin using vaginal samples collected from 49 women with clinically confirmed bacterial vaginosis. Gardnerella was detected in 42 of the 49 samples and was the dominant genus in 26.
When those samples were treated ex vivo, BNT331 endolysin reduced viable Gardnerella by at least 94% at concentrations of 20 to 50 µg/mL over 19 hours. The effect was sufficiently reproducible for the researchers to use the ex vivo work to inform dose selection for human testing. The microbiome effects were not completely restricted to Gardnerella: Lactobacillus iners also declined substantially, while Lactobacillus crispatus could continue proliferating where it was initially present in sufficient abundance. That nuance is important because a “precision” antibacterial does not necessarily mean activity against only a single species under every biological condition.
BioNTech subsequently moved BNT331 into a first-in-human study, making the programme one of the clearest examples of a major pharmaceutical company translating a phage-derived antibacterial enzyme into clinical testing.
BNT331-01, registered as NCT06469164, was a randomized, double-blind, placebo-controlled Phase 1 study conducted in the United States. It enrolled 102 participants and contained two parts. Healthy women received single ascending doses to evaluate safety and pharmacokinetics, while participants diagnosed with bacterial vaginosis received multiple ascending doses. In the BV part of the trial, treatment was administered through vaginal inserts over five consecutive days.
The study assessed safety, tolerability, pharmacokinetics and pharmacodynamics, with exploratory efficacy measurements in women with bacterial vaginosis. BioNTech now lists the trial as complete and provides scientific and plain-language result summaries through its clinical-trial portal. The study ran from July 2024 to July 2025, and registry results became publicly available in October 2026.
But the clinical milestone comes with an important corporate development that changes how BNT331 should be described today.
During the first quarter of 2026, BioNTech management decided to pause the BNT331 programme for what the company called strategic considerations. BioNTech consequently recognized a €29.9 million impairment associated with the candidate and reduced the recoverable value of the underlying intangible asset to zero. The wording matters: the company reported a decision to “pause” the programme rather than announcing that the scientific hypothesis had been disproven. Nevertheless, it means BNT331 should not currently be presented as an actively advancing clinical candidate.
The decision also occurred during a much broader strategic prioritization inside BioNTech. By 2026 the company was investing heavily in a large late-stage oncology portfolio, including multiple pivotal trials, while reporting cost reductions and lower expenditure associated with non-focus programmes. BNT331 disappeared from BioNTech's major 2026 corporate pipeline updates even though the clinical study remained registered as completed and the broader synthetic-lysin platform remained publicly described.
This creates an unusual situation. BioNTech has completed a 102-participant first-in-human study of a phage-derived antibacterial, published supporting ex vivo research and retained public descriptions of its lysin technology, yet its lead clinical lysin programme has been placed on strategic pause.
The underlying research effort, however, extends considerably beyond BNT331.
In August 2025, researchers entirely affiliated with BioNTech R&D Austria published a new screening method for identifying endolysins capable of killing Gram-negative bacteria. This is a substantially harder biological problem than targeting many Gram-positive organisms because Gram-negative bacteria possess an outer membrane that prevents most externally administered native endolysins from reaching the peptidoglycan layer they are designed to destroy.
The BioNTech team developed an agar-based screen that evaluates candidate enzymes against living Pseudomonas aeruginosa rather than only against purified or damaged bacterial material. From a library of engineered lysins, the researchers identified three candidates designated L0779, L0782 and L0799.
All three demonstrated bactericidal activity against P. aeruginosa under the experimental conditions, but L0782 emerged as the strongest candidate. It incorporates an outer-membrane-permeabilizing sequence designed to facilitate access to the peptidoglycan compartment. Against four tested P. aeruginosa strains, L0782 produced minimum inhibitory concentrations of 8–16 µg/mL. It also showed activity against Escherichia coli, Klebsiella pneumoniae and Acinetobacter baumannii, with MIC values reaching 4–16 µg/mL for several of these clinically important Gram-negative organisms.
These results remain preclinical and laboratory conditions can substantially influence apparent lysin activity. The authors themselves noted that buffer composition and membrane permeabilization can exaggerate the performance of some Gram-negative endolysins. The significance of the work lies less in having produced a new clinical candidate than in showing that BioNTech R&D Austria is developing a discovery platform for engineered phage enzymes extending well beyond bacterial vaginosis.
The publication is especially relevant to antimicrobial resistance because P. aeruginosa, K. pneumoniae and A. baumannii belong to bacterial groups for which treatment of multidrug-resistant infections can become extremely difficult. An engineered lysin capable of bypassing the Gram-negative outer membrane would potentially access a highly conserved structural target that conventional antibiotics do not exploit in the same way.
A second BioNTech R&D Austria programme became visible in 2026 and pursued a different engineering problem: how to transform highly active antibacterial enzymes into molecules suitable for systemic treatment.
Lysostaphin is a bacteriolytic enzyme with powerful activity against Staphylococcus aureus, but its potential as a drug has historically been constrained partly by short systemic persistence. BioNTech researchers therefore engineered constructs combining lysostaphin with an additional CHAP enzymatic domain and a human IgG1 Fc fragment intended to extend circulating half-life.
The resulting Fc-CHAP-LSN molecules retained strong antibacterial activity against S. aureus and coagulase-negative staphylococci. The researchers also reported activity in ex vivo S. aureus biofilm models and efficacy in mouse models of sepsis. The work was published in Frontiers in Microbiology in September 2026 by researchers including Michele Mutti, Manuel Zerbs, Lorenzo Corsini and Adriana Badarau from BioNTech R&D Austria.
Again, this is not evidence that BioNTech has opened another clinical programme. No human trial of these Fc-CHAP-LSN constructs has been announced. But the study demonstrates a second engineering direction within the same Vienna research operation: rather than simply finding naturally occurring phage enzymes, researchers are modifying antibacterial proteins to improve potency, broaden mechanisms or solve pharmacokinetic limitations.
Taken together, the Gardnerella, Gram-negative and Staphylococcus programmes reveal a broader technical strategy. BioNTech's antibacterial work is not centred on administering intact therapeutic phages. It is focused on extracting phage-derived antibacterial functions and treating them as engineerable biological drugs.
That distinction is important.
Whole-phage therapy benefits from viral replication at the site of a susceptible bacterial population, but it also inherits the biological complexity of an evolving virus–host interaction. Host range can be narrow, bacterial antiphage defence systems may prevent replication and the immune system can influence viral clearance. Manufacturing also requires propagation of each phage in a suitable bacterial host.
Endolysins behave more like conventional recombinant biologics. They do not need to infect the bacterium, reproduce inside it or evade intracellular antiphage immunity. Instead, their antibacterial effect depends on gaining access to the bacterial cell wall and cleaving specific peptidoglycan bonds.
They create different development challenges. Proteins can be cleared rapidly from circulation, neutralizing antibodies may develop after repeated exposure, Gram-negative outer membranes can prevent target access, and sufficiently broad biological activity must be balanced against the precision that makes the approach attractive in the first place.
BioNTech's programmes appear designed around exactly these limitations. BNT331 uses local administration for a microbiome-associated infection in which Gardnerella is readily accessible. L0782 incorporates membrane-permeabilizing functionality to attack Gram-negative bacteria. The Fc-CHAP-LSN constructs combine enzymatic domains with an Fc region to improve systemic pharmacological properties.
There are signs that BioNTech has considered an even more unusual extension of the concept: using mRNA to make the patient's own cells produce antibacterial lysins.
Patent applications associated with BioNTech describe mRNA constructs encoding peptidoglycan hydrolases, including phage-derived endolysins, with secretion signals intended to allow host cells to produce therapeutic antibacterial proteins. Named inventors include BioNTech researchers and executives such as Uğur Şahin, Lorenzo Corsini, Zehra Visram and Stephanie Erbar. A patent does not establish that such a product is being developed clinically, and BioNTech has not publicly announced a corresponding human programme, but the filings illustrate how the company's expertise in mRNA delivery could theoretically intersect with its lysin technology.
The concept would be markedly different from both conventional phage therapy and administration of purified recombinant endolysins. Instead of manufacturing the antibacterial protein as the final therapeutic product, an mRNA construct could encode it and use the patient's cells as transient production sites.
Whether this would offer a meaningful therapeutic advantage remains unknown. Expression level, tissue targeting, duration, immune responses and protein secretion would all need to be controlled. Nevertheless, the intellectual-property activity shows that BioNTech has explored antibacterial applications at the interface between two of its technological capabilities.
BioNTech is also not the only group pursuing Gardnerella-specific endolysins. Independent researchers associated with CC Biotech previously identified dozens of Gardnerella-active lysins and developed candidates including CCB7.1. Those studies demonstrated activity against Gardnerella in polymicrobial vaginal biofilms and reinforce the broader idea that phage-derived enzymes may offer a way to manipulate BV-associated bacterial communities more selectively than conventional antibiotics.
What distinguished BNT331 was how far BioNTech took the approach clinically. A 2026 review of phage-derived interventions for bacterial vaginosis identified BNT331 as the only registered intervention of its type that had reached an interventional human trial at the time of the authors' database search.
The programme therefore remains informative even if it does not immediately proceed into Phase 2.
BNT331 demonstrates that a recombinant phage-derived lysin can progress from phage biology through protein engineering, ex vivo microbiome experiments, formulation and finally randomized first-in-human testing inside one of the world's largest biotechnology companies.
The corporate decision to pause it also illustrates another reality of drug development: scientific feasibility and strategic priority are different questions.
BioNTech entered 2026 with a rapidly expanding oncology portfolio and substantial investment in late-stage assets. The €29.9 million BNT331 impairment occurred explicitly following a strategic decision to pause the programme, not an announcement that the entire LysinBuilder technology had failed. At the same time, BioNTech R&D Austria continued publishing antibacterial research against Gram-negative ESKAPE pathogens and Staphylococcus, while BioNTech's current website continues to describe synthetic lysins as part of its approach to bacterial infection and antimicrobial resistance.
That makes the future of this activity difficult to reduce to a simple “active” or “discontinued” label.
BNT331 itself is paused. Its Phase 1 trial is complete. The LysinBuilder platform remains publicly described. Researchers from BioNTech R&D Austria continue to generate peer-reviewed work on new antibacterial lysins. But BioNTech has not announced another clinical lysin candidate or a timetable for restarting BNT331.
For the phage field, the programme is nevertheless significant because it shows one route by which bacteriophage biology can enter mainstream pharmaceutical development without using intact phages at all.
Phages have spent billions of years evolving molecular systems capable of penetrating, manipulating and ultimately destroying bacterial cells. Therapeutic development can attempt to use the viruses themselves, but it can also isolate individual components of that machinery and engineer them as standalone drugs.
BioNTech chose the second route.
Its researchers have now explored a Gardnerella-selective endolysin in humans, enzymes capable of attacking Gram-negative ESKAPE organisms, engineered anti-staphylococcal constructs with prolonged systemic exposure and even intellectual-property concepts combining lysins with mRNA delivery.
None of those preclinical programmes should be interpreted as evidence that BioNTech is about to launch a new antibacterial product, and BNT331's strategic pause is an important warning against overstating the current commercial commitment.
But BioNTech's experience provides something unusual in a field dominated by specialist phage companies and academic laboratories: evidence that phage-derived antibacterial technology has been taken seriously enough by a major biotechnology company to support an acquisition potentially worth up to €150 million, build an internal research platform and advance one engineered lysin into a 102-participant first-in-human trial.
For a company whose public identity is dominated by mRNA and cancer immunotherapy, that makes BioNTech R&D Austria one of the more unexpected places to watch in the broader bacteriophage ecosystem.
The important question is no longer whether BioNTech ever entered phage-derived antibacterial development. It clearly did.
The question now is whether the science generated by PhagoMed, LysinBuilder and BioNTech R&D Austria will produce another clinical candidate after BNT331, be redirected into partnerships, or remain primarily as a research capability while the company concentrates its capital on oncology and other strategic priorities.
As of October 2026, BioNTech has not provided that answer.
Sources :
BioNTech / SEC — acquisition of PhagoMed Biopharma, October 2021. The acquisition included €50 million upfront consideration, less acquired debt, and up to €100 million in additional clinical-development milestone consideration.
https://www.sec.gov/Archives/edgar/data/1776985/000177698522000019/R14.htm
BioNTech — Protein-Based Platforms: LysinBuilder.
https://www.biontech.com/us/en/home/pipeline-and-products/platforms/protein-based-platforms.html
BioNTech — Infectious Diseases research and synthetic lysins.
https://www.biontech.com/us/en/home/research-and-innovation/therapeutic-areas/infectious-diseases.html
BioNTech — BNT331-01 clinical trial, NCT06469164.
https://clinicaltrials.biontech.com/trials/BNT331-01
ClinicalTrials.gov — NCT06469164.
https://clinicaltrials.gov/study/NCT06469164
BioNTech / SEC — First Quarter 2026 financial report. BioNTech states that management decided to pause BNT331 for strategic considerations and recorded a €29.9 million impairment.
https://www.sec.gov/Archives/edgar/data/1776985/000177698526000031/bntxq12026ex991quarterlyre.htm
BioNTech — Second Quarter 2026 corporate and financial update.
https://www.biontech.com/us/en/home/mediaroom/news/press-releases/2026/08/BioNTech-Announces-Second-Quarter-2026-Financial-Results-and-Corporate-Update.html
Landlinger C. et al. — Preclinical development of the Gardnerella-specific PM-477 endolysin. Antimicrobial Agents and Chemotherapy (2022).
https://doi.org/10.1128/aac.00264-22
Durica-Mitic S. et al. — “Endolysin selectively kills Gardnerella ex vivo in vaginal samples from women with bacterial vaginosis.” npj Biofilms and Microbiomes (2025).
https://www.nature.com/articles/s41522-025-00764-0
Oberbauer V. et al. — “Discovery of potent endolysins against Pseudomonas aeruginosa and other gram-negative ESKAPE pathogens by an agar-based screening method.” Scientific Reports 15, 29351 (2025).
https://doi.org/10.1038/s41598-025-14753-y
Mutti M. et al. — “A chimeric, half-life extended lysin with a unique mode of action.” Frontiers in Microbiology (2026).
https://doi.org/10.3389/fmicb.2026.1880134
BioNTech patent application — mRNA encoding peptidoglycan hydrolases and antibacterial proteins.
https://patents.google.com/patent/US20250027066A1/en
Review of the vaginal virome and phage-derived interventions for bacterial vaginosis, including the clinical-development status of BNT331. Microorganisms (2026).
https://www.mdpi.com/2076-2607/14/9/2055
.png){kind=logo}
Comments
Post a Comment