Adhesion Module

In the context of intranasal delivery, stable adhesion of probiotics is crucial for prolonging retention time and enhancing the delivery efficiency of therapeutic molecules. To this end, we designed a controllable adhesion module based on Lactobacillus plantarum WCFS1 and Escherichia coli Nissle 1917 (EcN). This design leverages both host cell-specific binding mechanisms and engineered optimization strategies to enhance the colonization and functional persistence of the engineered bacteria within the nasal cavity.

We plan to introduce the OppA protein and localize it extracellularly via a signal peptide to enhance its binding to the nasal epithelial-specific receptor (NaHS). In certain engineering schemes, a specific OppA protein subtype (Lp_0018) is screened and incorporated to optimize binding affinity.

Using a pSIP403-compatible vector to drive the expression of the target adhesion protein, WCFS1 expresses and secretes OppA under induction conditions, enabling binding to the nasal epithelial-specific receptor NaHS. Through engineered design, EcN can further function as a "scaffold," synergizing with WCFS1 to prolong its retention time and enhance colonization efficiency.

Fig.3 Targeted colonization facilitated by OppA protein screening

We have also developed an adhesion toolkit to achieve more robust bacterial adhesion. By leveraging antigen-antibody interactions, we established a dual-bacterial system comprising WCFS1 and EcN. This system utilizes synergistic adhesion and interaction mechanisms to enable enhanced and stable colonization.

Fig.4 Schematic Diagram of Dual-Bacteria Colonization

Therapeutic Module

In this system, the selected therapeutic factors are levodopa (L-DOPA) and glutathione, which exhibit synergistic effects in the treatment of Parkinson's disease:

Levodopa

Levodopa, a dopamine precursor, is continuously synthesized and released by the engineered bacteria. After absorption through the nasal mucosa, it can enter the brain via the olfactory nerve pathway or systemic circulation. There, it is decarboxylated by aromatic L-amino acid decarboxylase to yield the active therapeutic agent, dopamine. This process directly replenishes the cerebral dopamine deficit, thereby alleviating motor symptoms^[2].

The biosynthetic pathway for L-DOPA primarily relies on cellular L-tyrosine production. L-tyrosine can be hydroxylated to generate L-DOPA by 4-hydroxyphenylacetate 3-hydroxylase or tyrosinase^[3]. The gene cluster comprising hpaB and hpaC genes collectively expresses 4-hydroxyphenylacetate 3-hydroxylase. We heterologously expressed this gene cluster in Escherichia coli Nissle 1917 to enable L-DOPA production.

Fig.5 Metabolic Pathway of Levodopa

Glutathione

Glutathione, a crucial endogenous antioxidant, mitigates oxidative stress-induced damage to dopaminergic neurons and exerts neuroprotective effects. Dopamine metabolism itself generates substantial reactive oxygen species, leading to neuronal injury. Glutathione directly neutralizes free radicals and serves as a substrate for glutathione peroxidase to eliminate peroxides, thereby alleviating oxidative damage. Furthermore, glutathione depletion can initiate a vicious cycle, exacerbating mitochondrial dysfunction, neuroinflammation, and aberrant protein aggregation^[4]. Consequently, the intranasal delivery of glutathione via adherent Escherichia coli aims to directly replenish the impaired antioxidant defense system in the patient's brain, disrupt this detrimental cycle, provide neuroprotection at its root, and delay disease progression.

Glutathione biosynthesis is a two-step enzymatic process initiated by the precursor cysteine: First, glutamate-cysteine ligase, encoded by the gshA gene, catalyzes the conjugation of L-glutamate and cysteine to form γ-glutamylcysteine (γ-GC). Subsequently, glutathione synthase, encoded by the gshB gene, links γ-GC with glycine to ultimately generate glutathione (GSH). During various intracellular redox reactions, GSH is consumed and oxidized to form glutathione disulfide (GSSG). GSSG can then be regenerated into GSH by glutathione reductase, thereby maintaining cellular redox homeostasis^[5]. We expressed the gshA and gshB gene cluster in Escherichia coli Nissle 1917 to enable glutathione production.

Fig.6 Metabolic Pathway of Glutathione

Control Module

Controlled release has consistently been a pivotal challenge in drug delivery. We envisioned a system where therapeutic release from our microbial consortium is precisely regulated and deeply integrated with the adhesion module: therapeutic factors are produced only upon adhesion, and their release simultaneously acts to further enhance the adhesive capability.

This requirement brought "quorum sensing" (QS), a form of bacterial communication, into focus. We proposed to establish bidirectional QS between E. coli and L. plantarum to functionally couple their roles. This enables L. plantarum to trigger drug production in E. coli, while E. coli, in turn, enhances the adhesion capacity of L. plantarum.

Version 1.0: Communication based on AHLs

In our first version of the design, we aimed to establish bidirectional communication between Lactobacillus plantarum and Escherichia coli using AHL-based signaling:

Nissle 1917 was engineered to express the LuxI protein, enabling synthesis of a small, diffusible signaling molecule—30C6-HSL, an acyl-homoserine lactone (AHL). As cell density increases, 30C6-HSL accumulates both intra- and extracellularly in both strains. It binds to and activates the LuxR transcriptional regulator in L. plantarum, which in turn induces the expression of the OppA adhesion protein under the control of the Plux promoter, thereby enhancing the consortium's colonization capability in the olfactory epithelium.

Similarly, L. plantarum was engineered to express RhlI to produce C4-HSL. When C4-HSL diffuses into E. coli and reaches a threshold concentration, it binds to and activates the RhlR transcriptional regulator, leading to the induction of 4-hydroxyphenylacetate 3-hydroxylase (HpaBC) and bifunctional glutathione synthetase (GshAB) under the control of the Prhl promoter, thereby increasing the production of L-DOPA and glutathione.

Fig.7 Communications based on AHLs

However, our experimental results revealed that the native Plux promoter (BBa_R0062) exhibited no transcriptional activity in L. plantarum, regardless of the concentration of externally supplied 30C6-HSL, as detailed in the engineering section.

Literature indicates that promoter sequences differ significantly between Gram-positive and Gram-negative bacteria in terms of conservation, structure, and function ^[6-7]. Direct substitution often leads to promoter inactivation, which prompted us to consider promoter engineering.

Version 1.5: Engineering the Plux Promoter in Lactobacillus plantarum

Our promoter engineering strategy was adapted from the method developed by the Zeng team ^[8], which focused on modifying the Plux promoter in Bacillus subtilis. Similarly, we retained the Lux Box of the Plux promoter and replaced the -35 box, -10 box, spacer sequence, and transcriptional start region (TSR) with the corresponding consensus sequences for L. plantarum.

Fig.8 Engineering of the Plux Promoter

We conducted small-scale mutagenesis ourselves and also tested several improved promoter variants from the Zeng team. However, none of these yielded satisfactory results in our tests, as detailed in the engineering section.

We cannot rule out the overall feasibility of the Plux promoter engineering direction. However, constrained by time and experimental resources, we were unable to perform high-throughput mutagenesis and screening. Should subsequent teams pursue related research, they may consider excluding the specific variants we tested.

Consequently, we were compelled to strategically abandon the AHL-based cross-species communication approach and instead pivot to testing AIP-based cross-species communication.

Version 2.0: L. plantarum→E. coli: AHL E. coli→L. plantarum: AIP

In our final design, we retained the AHL-based communication from L. plantarum to E. coli, while establishing communication from E. coli to L. plantarum via AIP.

Among the various AIP systems, we selected the spp system due to its prior validation and well-characterized profile in L. plantarum. The primary challenge lay in enabling E. coli to secrete SppIP (a 19-amino acid peptide).

Since E. coli lacks a native AIP communication pathway and considering the fundamental differences in cell structure and secretion mechanisms between Gram-positive and Gram-negative bacteria, we opted not to directly employ the original SppIP secretion pathway (SppT/E). Instead, we leveraged the E. coli Type I Secretion System (T1SS). The T1SS is an efficient native mechanism in E. coli suitable for secreting short peptides under 100 amino acids ^[9].

Based on this system, we redesigned the communication scheme:

E. coli was engineered to express CvaA and CvaB proteins, which form a transmembrane complex with the TolC protein on the outer membrane. We fused the SppIP sequence to the CvaC15 signal peptide. This CvaC15 signal peptide directs SppIP to the CvaAB complex. After cleavage by CvaB, SppIP is transported extracellularly through the channel formed by the CvaA and TolC complex. The secreted SppIP then binds to the histidine kinase (SppK) on the cell membrane of L. plantarum. Subsequently, SppK undergoes autophosphorylation and transfers the phosphoryl group (~P) to the response regulator (SppR). The phosphorylated SppR then activates the Pspp promoter, driving the expression of the OPPA protein and thereby enhancing the colonization ability of L. plantarum.

Conversely, L. plantarum was engineered to express LuxI to synthesize 30C6-HSL. When 30C6-HSL diffuses into E. coli and reaches a threshold concentration, it binds to and activates the LuxR transcriptional regulator. This, in turn, induces the expression of 4-hydroxyphenylacetate 3-hydroxylase (HpaBC) and bifunctional glutathione synthetase (GshAB) under the control of the Plux promoter, ultimately increasing the production of L-DOPA and glutathione.

Fig.9 Achievement of Bidirectional Communication Between Lactobacillus plantarum and Escherichia coli

Safety Module

Considering the potential for colonizing bacteria to detach from the olfactory epithelium and escape into the external environment or other internal sites (particularly the intestinal tract), we have designed three distinct suicide switches to enhance the biosafety of our project. These include a thermosensitive suicide switch activated upon escape into the external environment, an anoxia-responsive suicide switch triggered in the intestinal environment, and an inducible suicide system for emergency containment. For detailed information, please refer to the Safety page.

Fig.10 Schematic diagram of the safety module