Introduction

Since the engineered bacteria cannot quickly gain a numerical advantage in the sewage environment, we hope that the AHL enzymes secreted by the engineered bacteria can more efficiently degrade multiple AHL compounds. Therefore, optimization of the wild-type AHL enzymes is required to achieve higher catalytic efficiency.

On the other hand, antimicrobial peptides(AMP) used to kill sulfate reducing bacteria(SRB) lack selectivity, and there is still host bacterial toxicity in direct secretion. Therefore, we hope to design a carrier targeting the SRB microenvironment in the form of a fusion protein,a chimeric antimicrobial peptide to achieve specific killing of SRB.

1. Semi-Rational Design of AHL Enzyme Optimization

The directed evolution of enzymes provides an effective means of obtaining advantageous mutants, but it often requires a significant amount of time and effort to construct a mutant library and carry out screening. On the other hand, semi-rational design involves saturated mutagenesis of a select few specific residues as targets based on the protein's structure-function relationship, greatly reducing the workload for screening.

1. Two AHL enzymes, AiiA and AidH

AiiA and AidH are both AHL lactonases, which catalyze the hydrolysis and ring-opening of AHL molecules, disrupting their activity as quorum sensing molecules. This disruption effectively interferes with quorum sensing in sulfate-reducing bacterial populations.

2. The structure of AiiA and AidH

We analyzed the structures of AiiA and AidH and found:

(1) AiiA belongs to a class of metal-dependent lactonases, where several histidine residues in the active pocket coordinate with metals, thereby conferring enzymatic activity.

(2) AidH is not a metal-dependent enzyme but rather a member of the α/β hydrolase enzyme family. (Mei, Gui-Ying et al. 2010) Based on multiple sequence alignments of AidH with its close relatives QqlG and QqlM, as well as literature research, it has been found that the catalytic center of AidH consists of a catalytic triad composed of three amino acids (S102, E219, H248). These three amino acids are highly conserved in the lactonase enzyme family.

(3) The catalytic mechanism of AidH involves the hydroxyl group of serine at position 102 acting as a nucleophilic agent, attacking the carbonyl carbon atom of the lactone. The lone pair electrons on the nitrogen of histidine at position 248 enhance the nucleophilicity of the serine hydroxyl group. Additionally, the carboxylate anion of glutamate at position 219 forms a hydrogen bond with the nitrogen atom on histidine, further increasing the electron density on the nitrogen atom. (Gao, Ang et al. 2013)

3. Mutation Site Analysis and Screening of AiiA

Since AiiA is a metalloenzyme, zinc ion coordination is a necessary condition to maintain its activity, following a general strategy of enzyme site-directed mutagenesis.(Ben Mabrouk, Sameh et al. 2011)

Considering introducing new salt bridges into the AiiA enzyme molecule is a normal way to enhance its stability. In accordance with the aforementioned mutation strategy, we selected to randomly mutate the 65th amino acid (asparagine) and the 195th amino acid (serine) of the AiiA protein and constructed a mutant library. As salt bridges need to be formed, we used basic amino acids to construct the mutant library.

We selected the following five mutants and then constructed three-dimensional models. For point mutations, the accuracy of homology modeling is already high enough. Therefore, we chose SWISS-Model to model our mutant proteins. We used PyMOL to analyze the RMSD values of the mutants and used CBdock2 software (Yang Liu, et al. 2022) to dock these five mutants with AHL molecules (C6-C12).

Combining the RMSD values and docking results, we found that point mutations did not affect the enzyme's active center, and the binding energy of each mutant with the substrate was similar to the wild type. Considering that the 70th amino acid, which is close in spatial position to the 65th amino acid of AiiA, is glutamic acid, and the 236th amino acid, which is close to the 195th amino acid of AiiA, is aspartic acid, both amino acids are acidic. To form salt bridges, we initially selected N65K and T195R as mutation candidates and entrusted the wet lab team to characterize and verify the activity of these mutants.

Fig.8 vital residues in AiiA(N65,E70,T195,D236)

4. Mutation Site Analysis and Screening of AidH

AidH belongs to the α/β hydrolase enzyme family. Based on the catalytic mechanism depicted in the diagram and the three-dimensional structure of AidH with the substrate, we have identified that the side chain of the 147th alanine residue might obstruct the binding of the substrate to the active pocket. (Zhang, Yixin et al. 2023) Therefore, we have chosen to mutate the 147th amino acid of AidH to alanine with a smaller steric hindrance and leucine with larger steric hindrance.

Based on the above, through semi-rational design, we have screened out 4 mutants: AiiA: N65K, T195R; AidH: A147G, A147V.

However, the success of the semi-rational design still requires validation through wet lab experiments. Therefore, we have entrusted the wet lab experimental team to assist us in constructing the mutant enzymes and verifying their activity.

2. Modeling of Enzyme-Antimicrobial Peptide Fusion Protein

1. Fusion protein of AMP and carrier

In many cases, directly expressing antimicrobial peptides(AMP) in Escherichia coli can be toxic to the host bacterium. (Yong Yang, Dejun Jin, Weiwei Yu, et al. 2017) Therefore, one common approach in prokaryotic expression systems is to produce antimicrobial peptides in the form of fusion proteins，because it didn’t inactivate AMP if its’ carbon terminal is extant. Additionally, since we do not want the freely released antimicrobial peptides to become toxic to the host bacteria after secretion, it is crucial to have a responsive antimicrobial peptide release strategy tailored to the specific microenvironment of the SRB.

2. Thiol-Based Biodegradable Linker Targeting SRB

We investigated the physicochemical characteristics of the SRB microenvironment and found that due to the extreme anaerobic conditions of SRB, the redox potential in its microenvironment is very low, and it generates numerous thiol compounds (Shi, Xuan et al. 2020) High concentrations of thiol compounds and an extremely low redox potential suggest that we can design a responsive antimicrobial peptide release strategy based on disulfide bond reduction. Therefore, we designed a degradable linker based on disulfide bonds as the fusion protein linker, enabling the release of antimicrobial peptides in response to the SRB microenvironment. This allows the antimicrobial peptides to accumulate in the SRB microenvironment, achieving a targeted effect.

3. Fusion protein design

Therefore, we selected the glycosidase DspB and AidH as the carrier of antimicrobial peptides because both enzymes are designed to secret into the extracellular,which means it can bring AMP as the same time, and designed the linker to be sensitive to TEV protease(which can recognize EQLYFQG). Therefore, the entire action process of our fusion protein is as follows:

1. E.coli expresses and secretes a fusion protein of Carrier-AMP, which is connected by peptide bonds and disulfide bonds.
2. E.coli expresses secreted TEV protease, which can efficiently recognize the sensitive linker between the two domains, hydrolyze the peptide bond, and leave only the disulfide bond between carrier and AMP
3. The disulfide bond is not easily broken in the microenvironment of E. coli, and is dispersed through DspB and AidH into the microenvironment of SRB. Due to the extreme anaerobic and reducing environment, the disulfide bond is opened and the antimicrobial peptide is released, achieving a rough targeting of SRB.

Therefore, the complete amino acid sequence is as follows:

DspB-SS Linker(with TEV recognition site)-Bactenecin(Bac):

MKKAITLFTLLCAVLLSFSTATYANAMDLPKKESGLTLDIARRFYTVDTIKQFIDTIHQAGGTFLHLHFSDHENYALESSYLEQREENATEKNGTYFNPKTNKPFLTYKQLNEIIYYAKERNIEIVPEVDSPNHMTAIFDLLTLKHGKEYVKGLKSPYIAEEIDINNPEAVEVIKTLIGEVIYIFGHSSRHFHIGGDEFSYAVENNHEFIRYVNTLNDFINSKGLITRVWNDGLIKNNLSELNKNIEITYWSYDGDAQAKEDIQYRREIRADLPELLANGFKVLNYNSYYLYFVPKSGSNIHNDGKYAAEDVLNNWTLGKWDGKNSSNHVQNTQNIIGSSLSIWGERSSALNEQTIQQASKNLLKAVIQKTNDPKSH GCP EQLYFQG CGILKR RLCRIVVIRVCR

AidH-SS Linker(with TEV recognition site)-Indolicidin(Ind):

MTINYHELETSHGRIAVRESEGEGAPLLMIHGNSSSGAIFAPQLEGEIGKKWRVIAPDLPGHGKSTDAIDPDRSYSMEGYADAMTEVMQQLGIADAVVFGWSLGGHIGIEMIARYPEMRGLMITGTPPVAREEVGQGFKSGPDMALAGQEIFSERDVESYARSTCGEPFEASLLDIVARTDGRARRIMFEKFGSGTGGNQRDIVAEAQLPIAVVNGRDEPFVELDFVSKVKFGNLWEGKTHVIDNAGHAPFREAPAEFDAYLARFIRDCTQ GC EQLYFQG CGILKR ILPWKWPWWPWRR

Among them, EQLYFQG is a sensitive sequence for TEV protease, which can be degraded by TEV protease.

4. Modeling of fusion protein

We used colabfold[10] to model our protein, The results indicate that the lDDT score of the antimicrobial peptide segment is very low, suggesting that the antimicrobial peptide segment is irregularly folded. However, what may actually affect the antimicrobial peptide activity is the post-cleavage structure, as there are additional amino acids on its N-terminus. Therefore, we calculated the RMSD value between the cleaved fused antimicrobial peptide and the original antimicrobial peptide. The results show that after cleavage, the structure of Bac remains essentially the same, indicating a greater likelihood of retaining activity. In contrast, the conformation of Ind undergoes significant changes, which may have a certain impact on its activity.

DspB-Bac

AidH-Ind

Fig.12 IDDT value of fusion proteins

Fig.13 Alignment of AMP and Carrier-AMP(digested by TEV protease)