Results

We converge on an integrated, probiotic-based platform that pairs immediate passive immunity via secreted cyclobodies (intein-cyclized nanobodies) with durable active immunity via surface display of conserved influenza antigens in Lactococcus lactis. This dual architecture simultaneously addresses the early outbreak protection gap and the need for immune memory, while keeping delivery oral, low-cost, thermostable, and field-ready.

Figure I. Dual-Immunization Platform constructs. a) Active Immunity for long-term protection, and b) Passive Immunity device for instant protection.

Passive arm (cyclobodies)

An inducible pLux/LuxR–AHL cassette with Npu DnaE split inteins and Usp45 secretion yields in vivo post-translational cyclization and export of nanobodies that are thermo- and protease-resistant, suitable for water/feed matrices. The cassette is modular (defined swap sites for Nano-A/Nano-B), enabling plug-and-play retargeting to new variants or even other viruses without rebuilding the backbone.

Active arm (L. lactis surface display)

Using PnisA (NICE) decouples growth from production. Usp45 + LPXTG anchoring is selected by epitope geometry/reach: SpaX for the compact LAH–4×M2e module and M6 for full-length NA. Multivalent M2e (4×, multispecies) and the orientation options (LAH↔4×M2e) aim to maximize epitope accessibility and immunogenic breadth, while avoiding interference with the passive arm's HA1/HA2 targets.

Final presentation/formulation

The end product is envisioned as a lyophilized blend containing (i) Lactococcus lactis expressing the surface antigens (active arm) and (ii) cyclobodies produced by the engineered chassis (passive arm). Lyophilization confers room-temperature stability, simplifies transport and storage, and enables rapid reconstitution in drinking water or feed, thereby minimizing cold-chain dependence and facilitating on-farm deployment.

What does the dual platform deliver?

Next steps (analytical, preclinical-ready)

Execute the in-vitro plan—FACS (quantify surface display of LAH, multivalent M2e, and NA on L. lactis), surface shaving (confirm true extracellular exposure and map accessible regions), fractionation (verify Usp45-mediated secretion and cell-envelope localization), ELISA / NA enzymatic activity (validate antigen integrity and cyclobody binding/neutralization proxies), and stability in water/feed (simulate farm matrices to define dose retention over time and handling constraints)—to confirm surface expression (active arm) and cyclobody robustness (passive arm). Once these acceptance criteria are met, advance to preclinical poultry studies focused on safety (tolerance, microbiome impact), dosing (CFU/mL in drinking water; cyclobody mg/L and re-dose interval), and performance under farm-relevant conditions (temperature/pH ranges, water quality, stocking density), using endpoints such as viral shedding, morbidity/mortality, weight gain, and feed conversion ratio (FCR).

In sum, this probiotic dual-immunization platform offers an accessible, adaptable, and operationally viable route to mitigate avian influenza (and other pathogens), combining instant passive protection with sustained active immunity, delivered as a lyophilized, orally reconstitutable product for on-farm use.

From the DBTL cycle in Engineered section, the results of AvianGuard for passive immunity showed that:

We engineered a passive-immunity platform that expresses and secretes cyclized nanobodies ("cyclobodies") using a split-intein cassette under pLux/LuxR–AHL control. The construct (Usp45sp → InteinC → CWN–G4S → Nano-A → linker → Nano-B → G4S–GGH → InteinN → Terminator) performs post-translational self-assembly, yielding a head-to-tail cyclic, bispecific nanobody that is more resistant to proteolysis and heat—well suited for field deployment in water/feed matrices. Strategic restriction sites flank Nano-A and Nano-B, allowing rapid swap-in of alternative binders for other targets or viral families.

What makes this platform stand out

The active immunity design was conceived with several components within the Lactococcus lactis. This is based on surface display of conserved viral antigens (influenza as proof-of-concept):

1. Antigenic Module A: LAH + 4×M2e

M2e

Origin: M2e corresponds to the N-terminal ectodomain (~23–24 amino acids) of the matrix protein 2 (M2) from avian influenza virus. It is a surface-exposed region involved in viral infectivity and is highly conserved across strains.

Immunological function: Its high conservation favors cross-reactive immune responses among influenza lineages.

LAH (Long Alpha Helix of HA)

Origin: LAH represents the long α-helix within the "stalk" domain of hemagglutinin (HA). This region, located at the base of HA, is structurally conserved and less prone to antigenic drift than the variable "head" domain.

Immunological function: Provides broad cross-protection; antibodies directed against the HA stalk can neutralize multiple influenza subtypes and strains.

2. Antigenic Module B: Neuraminidase (NA)

Origin: Neuraminidase (NA) is the second major glycoprotein on the influenza virus envelope (family Orthomyxoviridae).

Immunological function: Anti-NA antibodies inhibit sialidase activity, limiting viral spread and complementing the protection elicited by HA and M2e targets.

3. Secretion Signal (Usp45) and LPXTG Anchoring Domain

Usp45 (Sec): The standard secretion signal peptide in L. lactis that directs export of the polypeptide to the extracellular space.

SpaX: LPXTG-type covalent anchoring domain derived from Staphylococcal protein A (SpA) of Staphylococcus aureus, belonging to the Sortase A substrate family, enabling attachment to the peptidoglycan layer.

4. Expression System

Cp44: Strong constitutive promoter for L. lactis to ensure enough antigens production.

5. Linkers

Linkers used to minimize steric hindrance between the antigen and anchoring domain.

We designed a Lactococcus lactis surface-display platform for active immunity using conserved influenza antigens. The first design comprised two secreted, surface-anchored cassettes, both using an LPXTG anchoring domain (SpaX) under a strong constitutive promoter. The antigen layouts were:

The complete ORFs (including Usp45 and standard regulatory elements) were synthesized and cloned into expression vectors, and plasmids were propagated in E. coli cloning strains. Although synthesis and assembly were successful (Twist; ANSA), E. coli colonies carrying the constitutive cassettes were not recovered, mirroring the behavior seen in the passive-immunity constructs. The most parsimonious interpretation was that the strong constitutive promoter drove excessive basal expression during cloning, creating metabolic burden and membrane/periplasmic stress that impeded plasmid maintenance and colony recovery. This outcome underscored the need for temporal control of antigen expression to decouple growth from production.

To address this, we adopted a nisin-inducible promoter (PnisA, NICE system) to separate the growth and production phases. In parallel, we differentiated the anchoring domain by antigen while preserving the LPXTG motif, varying the stem architecture to tune exposure:

Using distinct LPXTG-type anchors enabled fine-tuning of protrusion and epitope accessibility, given that anchors in L. lactis differ in display efficiency and orientation on the cell wall (Michon et al., 2016; Plavec et al., 2019; Tay et al., 2021). Because NA is relatively large, an anchor with a longer/coiled-coil stem (e.g., M6/CwaM6) was expected to project the antigen farther from the surface, improving exposure—consistent with observations that coiled-coil scaffolds can extend proteins outward (Truebestein & Leonard, 2016). Conversely, the more compact LAH–4×M2e construct could benefit from SpaX, provided the linker length preserves accessibility and minimizes steric hindrance (Michon et al., 2016).

Note on apparent depth vs. exposure.

M6/CwaM6 typically contains extended coiled-coil stems that project cargo outward, whereas SpaX is more compact; thus, M6 generally protruded more and SpaX less. The actual presentation depended on linker length/flexibility and antigen size, so both configurations were planned with defined (G₄S)ₙ linkers and surface accessibility to be quantified by FACS or immunostaining (Ellis et al., 2023).

To broaden antigenic coverage and increase epitope density/avidity, we incorporated heterologous M2e variants. Specifically, we implemented a multistrain M2e design comprising four tandem M2e sequences—two avian, one swine, and one human—to address cross-species coverage and potential zoonotic relevance (Kim et al., 2012; Ding et al., 2021; Tan et al., 2021). In Module A, we retained the LAH domain within the same cassette and arranged the four distinct M2e sequences in tandem to maximize multivalency:

This approach leveraged two supported principles: (i) multimerization/tandemization of M2e markedly increased immunogenicity and breadth (Kim et al., 2012; Song et al., 2015; Gomes et al., 2022), and (ii) neuraminidase (NA)–specific antibodies inhibited sialidase activity and contributed independently to cross-protection, complementing stalk- and M2e-directed responses (Rockman et al., 2013; Abbadi et al., 2023). Prior Lactococcus lactis studies also demonstrated feasible surface display of NA or multimeric M2e (e.g., 4×M2e/10×M2e) and induction of mucosal/systemic immunity, supporting the chassis and the design logic (Reese et al., 2013; Lahiri et al., 2019). In sum, by enhancing valency and epitope density with multistrain M2e and pairing it with NA, the configuration was intended to improve cross-protective responses across multiple influenza lineages (Krammer, 2013; Kim et al., 2024).

After integrating multistrain M2e, we refined the spatial arrangement of epitopes within the fusion to optimize surface presentation and immune recognition. Because epitope orientation can influence accessibility, stability, and response quality, we compared alternative orders to determine which domain—LAH or M2e—should be positioned more externally in Module A. Two orders were defined for experimental comparison:

We posited that 4×M2e—LAH might enhance anti-M2e titers due to repetitive valency and high surface density, whereas LAH—4×M2e could favor exposure of conserved HA-stalk epitopes, potentially improving cross-clade breadth. Planned evaluation criteria included: surface accessibility by FACS or immunostaining, controlled surface-proteolysis assays, recognition by conformation-dependent antibodies, and monitoring of chassis load or stress responses. This structural fine-tuning emphasized balancing immunogenicity with expression stability to guide selection of the optimal antigen orientation for subsequent validation.

Our optimal design of the active-immunity architecture is a nisin-inducible, surface-display platform in L. lactis that exposes conserved viral epitopes on the cell wall while decoupling growth from expression (PnisA → Usp45 secretion → LPXTG anchoring). The system is built around interchangeable antigen modules so the same chassis can be rapidly re-targeted to different pathogens.

Extensibility to other viruses

The interchangeable modules accept any epitope or antigen domain that benefits from mucosal display. By swapping the inserts in A and/or B, the same platform can present, for example: coronavirus RBDs or S2-stalk segments, RSV F-protein stem peptides, henipavirus G fragments, conserved flavivirus E-stem regions, or parasite/bacterial adhesin epitopes. Selection guidelines:

Intended readouts

Induction tolerance and expression kinetics (growth curves under nisin), surface accessibility (FACS/immunostaining), protease shaving for topology, antigen-specific ELISA from mucosal/systemic samples, and functional assays (e.g., NA inhibition or pseudovirus neutralization), guiding per-antigen linker/anchor refinements.

To develop both the active and passive immunity modules of our project, we synthesized several DNA constructs through the companies Ansa Biotechnologies and Twist Bioscience.

1. Passive Immunity Module

Experimental Workflow:

The passive immunity module is associated with the production of cyclobodies—circular nanobody structures generated by intein-mediated protein splicing—the first experimental steps focused on verifying plasmid integrity and expression in Escherichia coli.

Upon receiving the synthetic DNA, we resuspended the lyophilized plasmids and transformed them into E. coli Top10 cells. Successfully transformed colonies were selected, and plasmid DNA was extracted for verification through restriction enzyme digestion.

2. Active Immunity Module

Experimental Workflow:

The strain L. lactis used for this system was obtained from our sponsor Cultivarium. Once verified as pure, the culture was propagated and preserved in MRS medium.

To prepare for transformation, L. lactis cells were made electrocompetent. Preliminary electroporation tests were conducted using a control plasmid to confirm cell competence and optimize voltage parameters.

After receiving the synthetic sequences, each construct was first transformed into E. coli Top10 to propagate and purify plasmids. Colonies were selected, and plasmid extraction was followed by restriction digestion to confirm sequence integrity. The next steps involve digestion and ligation of the synthesized epitopes (from Twist) into the active immunity backbone.

At the time of the wiki submission, experimental work is still in progress.

Once the epitopes are inserted and verified, their function will be assessed based on fluorescence emission (GFP/RFP) using a plate reader. Since this is a nisin-inducible system, adding nisin to the medium will activate expression, allowing us to evaluate the response and quantify activity through fluorescence intensity.

Because the arrival of synthetic parts and the L. lactis strain experienced delays, experimental progress was temporarily slowed, but results are expected soon as all essential components are now available.