Engineered Probiotics as Peptide Delivery Systems

Microbiome Peptide Profiling

GLP-1 produced directly in the gut by engineered bacteria

Genetically modified bacteria can colonize the intestine and continuously secrete therapeutic peptides, bypassing the injection requirement that limits most peptide drugs.

Brockmann et al., Science Advances, 2025

Brockmann et al., Science Advances, 2025

View as image

Peptide drugs face a fundamental delivery problem. Most cannot survive the digestive tract, so they require injection. Patients on GLP-1 receptor agonists inject weekly. Patients on insulin inject daily. Antimicrobial peptides degrade within minutes in gastrointestinal fluid. The entire oral peptide delivery field has spent decades trying to protect peptides from the gut environment long enough for absorption.

Engineered probiotics invert this logic. Instead of protecting a peptide from the gut, they produce the peptide inside the gut. Genetically modified bacteria colonize the intestinal tract and continuously secrete therapeutic peptides at the site where they are needed, or produce them for systemic absorption across the intestinal epithelium. The approach transforms a living organism into a programmable peptide factory that operates from within.

This concept sits at the intersection of synthetic biology, microbiome science, and peptide therapeutics. For the broader context of how gut microbes and peptide signaling interact, the pillar article on microbiome peptide profiling covers the diagnostic and mechanistic landscape. For how bacteria naturally produce antimicrobial peptides, see the sibling article on bacteriocin therapeutics.

The Platform: How Engineered Probiotics Produce Peptides

The engineering approach follows a common architecture across different bacterial chassis. A therapeutic peptide gene is cloned into a probiotic organism, placed under the control of a promoter (constitutive or inducible), and paired with a secretion signal that directs the finished peptide outside the bacterial cell. The probiotic colonizes the gut and continuously produces the peptide at the mucosal surface.

Romero-Luna et al. (2022) reviewed the genetic engineering and synthetic biology strategies used across food-grade bacteria. The key engineering decisions include: which bacterial species to use (safety profile, colonization ability, genetic tractability), which promoter system (constitutive for continuous production, inducible for controlled release), and which secretion pathway (direct secretion, surface display, or lysis-based release).^[1]

Chawla et al. (2026) provided the most comprehensive review of the field to date, categorizing recombinant live biotherapeutics by their delivery targets: anti-inflammatory cytokines (IL-10, IL-27), metabolic peptides (GLP-1, exendin-4), antimicrobial peptides, and immunomodulatory proteins. The review identified three generations of engineered probiotic design: first-generation (constitutive expression of a single peptide), second-generation (inducible expression with environmental sensors), and third-generation (multi-gene circuits with feedback control).^[2]

The three most commonly used chassis organisms are:

Escherichia coli Nissle 1917 (EcN): A probiotic E. coli strain with well-characterized genetics and strong colonization ability. EcN has the most sophisticated genetic toolkit available, supporting complex synthetic circuits with multiple inputs and outputs.

Lactococcus lactis: A food-grade lactic acid bacterium with a long safety history in dairy fermentation. L. lactis is the most clinically advanced platform, with the IL-10-secreting strain AG011 having completed phase I clinical trials for inflammatory bowel disease.

Saccharomyces boulardii: A probiotic yeast with natural gut colonization ability and eukaryotic protein processing machinery, enabling production of peptides that require post-translational modifications bacteria cannot perform. As a yeast, S. boulardii also resists antibiotics, avoiding interference when patients receive concurrent antibiotic therapy.

GLP-1 Delivery: Replacing the Injection

The most active research area is oral delivery of GLP-1 and related incretin peptides. Injectable GLP-1 receptor agonists (semaglutide, tirzepatide) are among the most prescribed drugs globally, but require weekly injection and cost over $1,000 per month. Engineered probiotics could produce GLP-1 continuously in the gut, eliminating the injection requirement.

Brockmann et al. (2025) engineered E. coli Nissle 1917 to secrete HldSE, a microbial peptide that locally elevates GLP-1 in the gut. In fiber-deprived mice (a model for the Western diet that depletes gut GLP-1 signaling), the engineered probiotic restored GLP-1 levels, normalized intestinal barrier function, and ameliorated colitis through GLP-1-dependent mechanisms. The approach worked by stimulating endogenous GLP-1 production from intestinal L-cells rather than delivering exogenous GLP-1 directly.^[3]

Hedin et al. (2023) took a direct delivery approach, engineering the probiotic yeast Saccharomyces boulardii to produce and secrete a GLP-1 receptor agonist (exendin-4) in the gastrointestinal tract. In combination with cold exposure (which activates brown adipose tissue), the engineered yeast produced anti-obesity effects in diet-induced obese mice. The yeast chassis offered advantages over bacterial platforms: eukaryotic protein folding machinery, natural resistance to antibiotics (avoiding interference with antibiotic therapy), and established safety as a probiotic.^[4]

Huang et al. (2026) demonstrated that recombinant Lactococcus lactis secreting GLP-1 analogs could restore pancreatic islet structure through the intestinal GLP-1 receptor pathway. The engineered bacteria delivered bioactive GLP-1 analogs directly to the intestinal mucosa, where GLP-1 receptors are abundant, achieving therapeutic effects through the natural incretin signaling axis rather than systemic drug exposure.^[5]

Ke et al. (2025) added a fourth chassis organism to the GLP-1 delivery landscape by engineering Lactobacillus gasseri to secrete a GLP-1 fusion peptide. The L. gasseri platform offered distinct advantages: strong native colonization of the human gastrointestinal tract, resistance to bile salts, and a well-characterized safety profile from decades of use in fermented foods. The recombinant strain produced bioactive GLP-1 that improved glycemic control in diabetic mouse models, demonstrating that multiple probiotic species can serve as effective GLP-1 delivery vehicles.^[11]

The implications for peptide therapeutics are substantial. For context on how oral peptide drug delivery is advancing through other approaches, see the future of oral peptide drugs.

Oral-to-Systemic Peptide Absorption

A critical question for engineered probiotic peptide delivery is whether peptides produced in the gut can reach the systemic circulation. For gut-local applications (inflammatory bowel disease, H. pylori infection), mucosal delivery is sufficient. For systemic targets (arthritis, metabolic disease), the peptide must cross the intestinal epithelium.

Kady et al. (2025) demonstrated oral-to-systemic delivery using engineered Limosilactobacillus reuteri that secreted iberiotoxin, a potassium channel blocker peptide. The probiotic delivered iberiotoxin across the gut barrier in sufficient quantities to treat collagen-induced arthritis in rats, an autoimmune model requiring systemic drug exposure. This was the first demonstration that an engineered probiotic could achieve therapeutic systemic peptide levels through oral administration.^[6]

Gelli et al. (2025) quantified the absorption mechanism using engineered Saccharomyces boulardii in the murine gastrointestinal tract. They showed that the probiotic yeast enhanced intestinal absorption of macromolecules through a combination of increased paracellular transport (between cells) and transcytosis (through cells). The enhancement was sufficient to deliver therapeutic macromolecule levels to the bloodstream from oral administration alone.^[7]

These findings address one of the central challenges in oral peptide drug development: achieving adequate bioavailability without chemical modification of the peptide itself.

Anti-Inflammatory Peptide Production

Beyond metabolic peptides, engineered probiotics can produce anti-inflammatory proteins directly at inflamed mucosal surfaces. The most clinically advanced example is Lactococcus lactis engineered to secrete interleukin-10 (IL-10), an anti-inflammatory cytokine. The strain AG011 completed a phase I clinical trial for Crohn's disease, demonstrating that locally produced IL-10 in the intestine reduced inflammation without the systemic side effects associated with intravenous IL-10 administration.

The local production model is critical for mucosal inflammation. Systemic IL-10 causes dose-limiting immunosuppression at concentrations needed for gut anti-inflammatory effects. An engineered probiotic producing IL-10 directly at the inflamed mucosa achieves high local concentrations with minimal systemic exposure. The same principle applies to other anti-inflammatory peptides and proteins: local production avoids the therapeutic index problems that plague systemic administration.

Skrlec et al. (2018) demonstrated another therapeutic application by engineering Lactococcus lactis to produce and secrete BPC-157, a cytoprotective pentadecapeptide with documented wound-healing and anti-inflammatory properties in animal models. The recombinant L. lactis secreted functional BPC-157 that retained antioxidant activity, demonstrating that the bacterial expression system preserved the peptide's biological function. The study established L. lactis as a viable oral delivery vehicle for gastroprotective peptides, a direct application of probiotic engineering to a peptide with clinical relevance in gut healing.^[10] For more on BPC-157's broader research profile, see BPC-157: The Body Protection Compound.

Targeted Antimicrobial Peptide Delivery

Engineered probiotics can produce antimicrobial peptides precisely at the site of infection, achieving local concentrations that would be impossible to reach through systemic administration while sparing beneficial bacteria elsewhere in the gut.

Choudhury et al. (2025) engineered probiotics that secreted guided antimicrobial peptides designed to selectively kill Fusobacterium nucleatum, a pathogen linked to colorectal cancer progression, inflammatory bowel disease, and periodontal disease. The engineered bacteria detected specific environmental signals associated with Fusobacterium infection and activated antimicrobial peptide production in response, creating a precision antimicrobial that operated only when and where needed.^[8]

Hu et al. (2026) engineered Lactococcus lactis to express antimicrobial peptide HI, demonstrating enhanced bacterial survival in the gut environment and protection against enterotoxigenic E. coli (ETEC) infection in mice. The engineered probiotic both colonized the gut and produced antimicrobial peptides, combining two therapeutic functions in a single organism.^[9]

Mejia-Pitta et al. (2021) provided a comprehensive review of the engineering strategies used across probiotic species for antimicrobial peptide production, cataloging expression systems (constitutive vs. inducible promoters), secretion pathways (Sec-dependent, Sec-independent, and lysis cassettes), and peptide stabilization approaches (fusion partners, protease-resistant modifications). The review noted that probiotic-delivered antimicrobial peptides face lower resistance development risk than conventional antibiotics because the peptides act through membrane disruption rather than metabolic pathway inhibition, and the local delivery concentrations achieved exceed resistance-selection thresholds.^[12]

This approach intersects with the broader field of antimicrobial peptides and microbiome balance. One of the persistent challenges with antimicrobial peptide therapy is collateral damage to beneficial gut bacteria. Engineered probiotics address this through two mechanisms: producing antimicrobial peptides locally (reducing exposure to distant microbiome communities) and incorporating target-recognition domains that guide the peptides to specific pathogens.

Engineering Challenges and Limitations

Colonization stability. Engineered probiotics must colonize the gut long enough to deliver therapeutic peptide levels. Colonization duration varies by species, diet, existing microbiome composition, and host immune status. Some strains persist for days; others wash through in hours. Repeated dosing may be required, which reduces the convenience advantage over conventional drugs.

Genetic stability. Bacteria evolve rapidly. Under selective pressure, engineered bacteria can lose their therapeutic gene cassettes (which impose a metabolic burden) within days to weeks. Genetic containment strategies (auxotrophies, kill switches, essential gene dependencies) can mitigate this but add complexity and potential failure points.

Dosing precision. Unlike a pill with a defined dose, an engineered probiotic produces peptides at a rate determined by bacterial population size, growth conditions, and gene expression levels, all of which vary between individuals and over time within the same individual. Achieving consistent therapeutic peptide levels is more challenging than with conventional dosing.

Regulatory pathway. Engineered probiotics are classified as live biotherapeutic products (LBPs) by the FDA and face a regulatory pathway distinct from both drugs and biologics. The path to approval is less established, with fewer precedents and more uncertainty about required safety and efficacy data. The most advanced candidate, AG011 (IL-10-secreting L. lactis for Crohn's disease), completed phase I trials but the program was discontinued.

Immune response. The host immune system actively surveys gut bacteria. Engineered probiotics producing foreign peptides may trigger immune responses that eliminate the engineered strain or cause inflammation. This is particularly relevant for peptides not normally found in the gut environment.

Peptide folding. Bacterial expression systems cannot perform all the post-translational modifications required for some therapeutic peptides. Disulfide bond formation, glycosylation, and other modifications may be absent or aberrant in bacterially produced peptides. Yeast platforms (S. boulardii) address some of these limitations through eukaryotic processing machinery.

These challenges are not unique to engineered probiotics. Many parallel the difficulties encountered in oral peptide drug development broadly, where achieving consistent bioavailability remains the central unsolved problem. The engineered probiotic approach sidesteps the absorption barrier by producing the peptide on the mucosal side, but introduces its own set of variability sources (bacterial population dynamics, gene expression stability) that conventional formulations do not face.

For how the gut-brain peptide signaling axis intersects with these delivery approaches, see the microbiome-peptide axis and gut peptide hormones.

The Bottom Line

Engineered probiotics convert living bacteria into programmable peptide factories that operate from within the gut. The three main chassis organisms (E. coli Nissle 1917, Lactococcus lactis, and Saccharomyces boulardii) have been engineered to produce GLP-1 analogs for metabolic disease, antimicrobial peptides for targeted pathogen killing, and anti-inflammatory cytokines for inflammatory bowel disease. Recent studies demonstrate that probiotic-produced peptides can achieve both mucosal and systemic therapeutic levels. The field faces practical challenges in colonization stability, genetic maintenance, dosing precision, and regulatory classification, but the fundamental proof of concept is established across multiple peptide types and disease models.

Frequently Asked Questions

Can engineered probiotics deliver peptide drugs orally?

Yes. Engineered bacteria and yeast can be programmed to produce therapeutic peptides directly in the gut after oral administration. Kady et al. (2025) demonstrated oral-to-systemic delivery of a peptide toxin using engineered L. reuteri, achieving therapeutic levels sufficient to treat arthritis in rats without injection. Multiple groups have delivered GLP-1 analogs, antimicrobial peptides, and anti-inflammatory cytokines through engineered probiotics.

What peptides can engineered probiotics produce?

The range includes GLP-1 and incretin analogs for metabolic disease (Brockmann et al., 2025; Hedin et al., 2023), antimicrobial peptides for targeted pathogen killing (Choudhury et al., 2025; Hu et al., 2026), anti-inflammatory cytokines like IL-10 for inflammatory bowel disease, and channel-blocking peptides for autoimmune conditions. The limiting factor is the bacterial cell's ability to correctly fold and secrete the target peptide.

Are engineered probiotics safe to consume?

The chassis organisms used (E. coli Nissle 1917, Lactococcus lactis, S. boulardii) have established safety profiles as probiotics. L. lactis engineered to secrete IL-10 completed a phase I clinical trial for Crohn's disease without safety concerns. Genetic containment strategies (kill switches, auxotrophies) prevent environmental release of engineered genes. Long-term safety data for peptide-producing strains remains limited, as most studies are preclinical.

Could engineered probiotics replace GLP-1 injections?

In theory, engineered bacteria that continuously produce GLP-1 in the gut could replace weekly injections. Hedin et al. (2023) showed anti-obesity effects from yeast-produced GLP-1R agonists, and Huang et al. (2026) restored pancreatic islet structure with L. lactis-produced GLP-1 analogs. Practical challenges remain: achieving consistent dosing, maintaining bacterial colonization, and clearing regulatory hurdles for live biotherapeutic products.

How do engineered probiotics target specific bacteria?

Engineered probiotics can be programmed with environmental sensors that detect pathogen-specific signals (quorum-sensing molecules, pH changes, metabolic byproducts) and activate antimicrobial peptide production only in response. Choudhury et al. (2025) demonstrated this with probiotics that selectively killed Fusobacterium nucleatum using guided antimicrobial peptides, sparing beneficial gut bacteria through molecular targeting.

What bacteria are used to make engineered probiotics?

Three organisms dominate the field. E. coli Nissle 1917 offers the most sophisticated genetic toolkit and strong gut colonization. Lactococcus lactis has the longest safety record from dairy use and is the most clinically advanced platform. The yeast Saccharomyces boulardii provides eukaryotic protein processing capabilities that bacteria lack, enabling production of peptides requiring complex folding or modifications.