A Chirality‐Converted Bacteriolytic Dodecapeptide Regulates Vibrio‐Induced Polymicrobial Infection and Ameliorates Invasion‐Associated Gut Microbiota Disequilibrium Authors: Ping Zeng, Qipeng Cheng, Xiaoxu Zhang, Honglan Wang, Jinghan Zhang, Xinyi Ding, Pengfei Zhang, Lanhua Yi, Kwok‐Yin Wong, Kin‐Fai Chan, Sheng Chen, Sharon Shui Yee Leung Published: 2026 ABSTRACT ABSTRACTThe polymicrobial infections caused by Vibrio species, especially multidrug‐resistant isolates, are posing increasing threats to the health of coastal residents. Herein, a novel dodecapeptide, denoted as D‐zp37, was designed and synthesized, aiming to combat these notorious pathogens. Experimental results proved that the simple chirality conversion of the parent peptide, zp37, could significantly boost the antibacterial activity and proteolytic stability. To a specific cephalosporin‐resistant Vibrio alginolyticus, the minimal inhibitory concentration value of D‐zp37 was as low as 0.5 µm. Beyond the strong bacteriolytic effect against planktonic Vibrio alginolyticus, Vibrio parahaemolyticus and Vibrio vulnificus strains, D‐zp37 was found to restrain the biofilm establishment of Vibrio mixtures via inhibiting the efflux of intracellular polysaccharides. Mechanistic studies hinted that D‐zp37 impeded the electron transport chain, impaired the membrane stress response by downregulating the phage shock protein family, and blocked amino acid biosynthesis in Vibrio cells. Furthermore, three different infection models (shrimp, Galleria mellonella larvae and immunosuppressed mice) collaboratively confirmed the anti‐Vibrio efficacy of D‐zp37 in vivo. In addition, the intragastric administration of D‐zp37 was found to be beneficial in restoring the imbalanced gut microbiota induced by the polymicrobial Vibrio invasion. In summary, this work investigated the antibacterial mechanisms and in vivo efficacy of a novel peptide lead compound, proposing a new therapeutic against Vibrio infections.Keywords: antibiofilm, cationic amphiphile, dextrorotatory amino acid, motility inhibition, mouse fecal microbiota, phage shock protein Introduction Vibrio is a genus of Gram‐negative (G‒ve) bacteria widely distributed in coastal waters and inside bodies of marine animals and many of them are aquatic pathogens [1]. Vibrio infections can lead to production loss of seafood and threaten the safety of final consumers. For human beings, Vibrio invasion can be acquired via intaking contaminated seafood or exposing open wounds to bacteria‐rich environments. The resulting vibrioses, represented by necrotizing fasciitis and watery diarrhea, pose potential health threats, especially to immunocompromised coastal residents [2]. Although great efforts have been put into fighting against Vibrio infections, hitherto they remain a thorny issue [3]. It is predicted that the population at risk of Vibrio infections would reach up to 1.3 billion by 2050 [4]. Clinically, commonly used antibiotics against Vibrio species include tetracycline, gentamicin, fluoroquinolones and β‐lactam antibiotics [5]. However, the abuse of these medications and the dissemination of plasmid‐mediated resistance greatly impair their effectiveness in controlling Vibrio infections [6]. Isolation of multidrug‐resistant (MDR) Vibrio alginolyticus (VA), Vibrio parahaemolyticus (VP) and Vibrio vulnificus (VV) were frequently reported [7]. Occurrence of tetracycline resistance gene tet(M) was first detected in Vibrio species about 20 years ago, while the novel metallo‐β‐lactamase VAM‐1, which conferred reduced carbapenem susceptibility, was identified more recently, highlighting the evolving antimicrobial resistance in this genus [8]. The constant warming marine environment is also believed to intensify the global expansion of MDR Vibrio [9]. Mature biofilms formed by Vibrio species can persist on the surfaces of commonly consumed seafoods [10]. This dense and complex extracellular matrix impedes the interaction between embedded bacterial cells and antibacterial agents, thus helping pathogens escape external antibacterial attacks [11]. Furthermore, in real‐world environments, diverse microorganisms can build cooperative communities to promote co‐existence [12]. Their spatial organization within wound sites can result in refractory polymicrobial infections [13]. Antimicrobial peptides (AMPs), originated from the host defense system, are considered promising alternatives in combating pathogens, especially in this antimicrobial resistance era. [14] The outer membrane (OM) of G‒ve Vibrio species presents a significant permeability barrier to many small‐molecule antibiotics [15]. In the meantime, the large amount of negatively charged lipopolysaccharide (LPS) on the OM surface could serve as important targets for cationic AMPs through electrostatic interactions. Therefore, designing LPS‐targeting peptides becomes a simple yet effective strategy for targeting G‒ve bacteria [16]. Previously, we identified a peptide zp3 (GIIAGIIIKIKK‐NH2) from a fragment of the Staphylococcus aureus toxin phenol‐soluble modulin, which showed antibacterial activity against both Escherichia coli and Bacillus subtilis [17]. Through targeted amino acid substitutions, its analog zp37 (GIKAKIIIKIKK‐NH2) exhibited an expanded antibacterial spectrum and improved bioactivity [18]. However, the sequence of zp37 consists entirely of proteinogenic amino acids, resulting in poor proteolytic stability, which significantly limits its clinical potential. In this study, 11 residues of zp37, except the capping glycine from the N‐terminal, were replaced with their corresponding dextrorotatory chiral isomers. This novel compound, denoted as D‐zp37, was investigated for its antibacterial capability against polymicrobial Vibrio infections using both in vitro and in vivo models. Results Chirality Conversion Enhances In Vitro Stability and Antibacterial Potential Against MDR Vibrio Strains of Designed Dodecapeptide Peptide zp37 was predicted to be sensitive to many common enzymes, such as thermolysin, proteinase K, and trypsin. Possible protease cleavage sites were visualized (Figure 1A(a)). Its helical structure and amphiphilic interface were shown in Figure 1A(b,c) respectively. To improve the proteolytic stability of zp37, its chirality‐converted analogue was engineered by replacing all L‐isoleucine (2S, 3S) with D‐isoleucine (2R, 3R). This novel peptide was denoted as D‐zp37. Chemical structures of zp37 and D‐zp37 were shown in Figure 1B(a,b), respectively. (A) In silico predicted possible protease cleavage sites (a) of zp37. Ribbon diagram (b) indicates secondary structure and surface diagram (c) indicates Van der Waals surface based on hydrophobicity. (B) Chemical structures of zp37 (a) and D‐zp37 (b). (C) MIC values of zp37 and D‐zp37 against Vibrio strains. (D) HPLC diagrams of zp37 (a) and D‐zp37 (b) after incubating with 10 µg/mL trypsin. Remaining peak areas (c) of zp37 and D‐zp37 after 120 min treatment. Retention time (d) of zp37 and D‐zp37. (E) Bacterial growth in various conditions in the presence of different concentrations of D‐zp37. (F) Live bacterial counting after D‐zp37 treatment. We first evaluated the antibacterial activity of the original and modified peptides against Vibrio bacteria using a double‐dilution assay. For the 15 tested MDR Vibrio isolates (5 VA, 5 VP and 5 VV strains), the minimal inhibitory concentration (MIC) values of D‐zp37 were at least 4‐fold lower than that of the parent peptide, zp37, and in the range of 0.5 to 8 µm (Figure 1C). The primary objective of engineering zp37 to D‐zp37 is to improve the proteolytic stability. We therefore compared their trypsin tolerance using high performance liquid chromatography (HPLC) to quantify the rate of degradation. Upon the change of chirality of isoleucine in the peptide sequence, the retention time of zp37 (6.8 min) was shifted to 5.8 min for D‐zp37 (Figure 1D(d)), which was common among chiral enantiomers [19]. The chromatography results revealed that zp37 degraded significantly within 30 min in the presence of 10 µg/mL trypsin (Figure 1D(a)). Differently, the same concentration of trypsin would hardly cause degradation of D‐zp37 throughout the entire monitoring period (Figure 1D(b)). After 120 min incubation, the remaining peak areas of zp37 and D‐zp37 were 35.92% and 96.74%, respectively (Figure 1D(c)). To further evaluate the environmental tolerance of D‐zp37, its bacterial inhibition effect was measured in various harsh conditions. We selected one strain from each of the three Vibrio species as indicators, namely VA 1579, VP 1647 and VV 5043. Figure 1E depicted the level of bacterial growth after incubating in normal medium, trypsin, fetal bovine serum (FBS), gastric juice (GJ), intestinal juice (IJ), and 3.5% NaCl solution in the presence of different concentrations of D‐zp37 for 6 h. All three Vibrio cells can be at least partially inhibited in the presence of 2 µm D‐zp37 in all tested scenarios. Figure 1F revealed that 20 min treatment with D‐zp37 at the respective MIC‐level was able to significantly reduce the colony forming unit (CFU) for all Vibrio species. Moreover, D‐zp37 at 2‐fold MICs could eradicate VA 1579 and VP 1647 to undetectable levels in 20 min. D‐zp37 Demonstrates Selectively Bacteriolytic Effect Against Vibrio Cells Co‐existence of various pathogens gives rise to polymicrobial infection [20]. The isolation of multiple Vibrio species from the same aquatic animal sample as primary colonizers has also been reported [21]. Therefore, a bacterial mixture composed of an equivalent amount of VA 1579, VP 1647 and VV 5043 was set as a model to evaluate the efficacy of D‐zp37 against Vibrio polymicrobial infection. First, the turbidity assay suggested that D‐zp37 exhibited bacteriolytic effect on the Vibrio mixture (Figure 2A). The optical density at 600 nm (OD600) values dropped significantly after 15 min peptide treatment, indicating that D‐zp37 was able to lyse the Vibrio mixture rapidly. Comparatively, antibiotics ceftazidime (CAZ) exhibited no killing effect at the tested concentration. Scanning electron microscopy (SEM) images described a portrait of Vibrio cells treated with phosphate buffered saline (PBS), 2 or 8 µm D‐zp37 respectively (Figure 2B). The morphology of bacterial cells became collapsed and fused after the peptide treatment. The loss of a well‐defined boundary clarified that Vibrio cells were being lysed by D‐zp37. While for red blood cells (RGB), D‐zp37 showed a negligible hemolytic effect even at 256 µm (Figure 2C). (A) Turbidity changes of Vibrio mixture after D‐zp37 or 8 µm CAZ treatment. (B) SEM images of Vibrio mixture after PBS (a), 2 µm D‐zp37 and 8 µm D‐zp37 treatment. Scale bar: 5 µm. (C) Hemolytic activity of D‐zp37 against RBCs. (D) Triple fluorescence staining of Vibrio mixture by FITC‐D‐zp37, DAPI and PI. Scale bar: 100 µm. (E) TEM images of the Vibrio mixture. Scale bar: 200 nm. (F) Cytotoxicity of D‐zp37 against GES‐1 cells. (G) Co‐culture system of Vibrio mixture and GES‐1 cells stained by Nile red, Hoechst 33258 and FITC‐D‐zp37. Scale bar: 100 µm. We further investigated the intracellular penetration kinetics of the peptide using fluorescein isothiocyanate‐labelled D‐zp37 (FITC‐D‐zp37) and fluorescence microscopy (Figure 2D). According to the images of the first column, the green fluorescence emitted from Vibrio cells could be observed in 15 min. With time elapsing, the fluorescence intensity increased significantly, suggesting that a growing amount of D‐zp37 molecules had crossed the cell membrane. On the contrary, blue fluorescence from 4’,6‐diamidino‐2‐phenylindole (DAPI), a DNA specific probe, gradually weakened, which hinted that DNA leakage may happen during bacterial lysis. Propidium iodide (PI) staining reconfirmed the bactericidal effect of D‐zp37. Fine structural changes of Vibrio after being attacked by D‐zp37 were captured by transmission electron microscopy (TEM) (Figure 2E). Bacterial cells in the PBS‐treated control group presented a healthy state of growth. In contrast, bacteria subjected to D‐zp37 treatment exhibited anomalous morphology including membrane exfoliation (2 µm group), cell deformation and cytoplasm leakage (4 µm group), and severe structural distortion and even complete bacterial lysis (8 µm group). The digestive tract is a high incidence site for Vibrio polymicrobial infection. To assess whether this molecule adversely affected normal host cells, we tested its cytotoxicity to the human gastric epithelial cell line GES‐1 (Figure 2F). Cell counting kit‐8 (CCK‐8) assay reported the half‐maximal inhibitory concentration (IC50) of D‐zp37 against GES‐1 was >64 µm, much higher than the MICs against Vibrio bacteria. Moreover, when a Vibrio mixture was co‐cultured with GES‐1 cells, D‐zp37 demonstrated a stronger affinity to bacterial cells. This was evidenced by fluorescence microscopy, where green fluorescence emitted by FITC‐D‐zp37 was predominantly observed within bacterial cells (stained in red) rather than mammalian cells (stained in blue) (Figure 2G). D‐zp37 Restrains Bacterial Motility and Combats Biofilm Establishment via Inhibiting Efflux of Intracellular Polysaccharide The flagella of Vibrio bacteria could boost motility, thereby facilitating their fast spread [22]. Swimming behaviour, as one of the representative forms of motion, is often monitored to evaluate whether an antimicrobial agent can suppress cluster migration of bacterial cells. For six consecutive days, we recorded the swimming images of Vibrio mixture on the surface of an agar plate containing different concentrations of D‐zp37 (Figure 3A). In general, the bacterial swimming area of ​​each group gradually expanded over time. However, in parallel comparison, the swimming area decreased with increasing D‐zp37 concentration, suggesting that the peptide could inhibit the migration of Vibrio cells in a dose‐dependent manner. Average diameters of 0 µm and 32 µm groups shown in Figure 3B were 3.58 and 2.35 cm on Day 1. The difference between the two groups was 1.23 cm. While on Day 6, such value widened to 1.97 cm (5.93 and 3.96 cm respectively). (A) Swimming behaviour of Vibrio mixture in agar plates containing different concentrations of D‐zp37. Scale bar: 2 cm. (B) Average swimming diameters of Vibrio mixture. (C) Fluorescence images of exopolysaccharide expressed by Vibrio mixture after PBS (a), or 2 µm (b), 8 µm (c), 16 µm (d) D‐zp37 treatment. Scale bar: 100 µm. (D) Quantitative analysis of exopolysaccharide content before and after D‐zp37 treatment. (E) Turbidity changes of Vibrio mixture and corresponding intracellular polysaccharide content before and after D‐zp37 treatment. (F) Productive biomass of biofilm in the absence or presence of D‐zp37. (G) Phase images of biofilm treated with PBS (a), or 16 µm (b), 32 µm (c), 64 µm (d) D‐zp37. Scale bar: 100 µm. Producing biofilm is an effective strategy for bacteria to resist antibiotics, necessitating the antibiofilm capability of AMPs for effective control [23]. Though the compositions of biofilms are complex, exopolysaccharide (EPS) are considered to be the core component [24]. The secretion amount of EPS after the D‐zp37 treatment was then visualized (Figure 3C) and quantitatively analyzed (Figure 3D). Fluorescence intensity of Alexa Fluor 488 conjugated concanavalin A decreased significantly when D‐zp37 was applied, suggesting that Vibrio mixture released less EPS in the presence of D‐zp37. Interestingly, when we tested the intracellular polysaccharide of the Vibrio mixture, it can be found that its content continued to increase despite sharply dropping bacterial turbidity (Figure 3E). After normalization, intracellular polysaccharide contents per unit cell rose 1.05‐fold (1 µm), 1.25‐fold (2 µm), 2.13‐fold (4 µm) and 6.53‐fold (8 µm) compared with the untreated control, respectively. As mentioned before, EPS play a key role in forming the biofilm matrix. The diminishment of EPS after D‐zp37 treatment was also reflected in the decrease of biomass of the biofilm (Figure 3F). Treating Vibrio mixture with 2 µm D‐zp37 could have a dramatical impact on biofilm establishment and bacterial cells incubated with >8 µm D‐zp37 basically lost the ability to produce biofilm. Moreover, D‐zp37 was able to dissipate pre‐formed biofilms (Figure 3G), Phase contrast images captured the difference of Vibrio biofilm treated with PBS, 16, 32 and 64 µm D‐zp37. Pre‐formed biofilm presented a dense and uniform structure (a of Figure 3G). However, D‐zp37 molecules triggered inhomogeneity in a dose‐dependent manner (b–d of Figure 3G), indicating that parts of the biofilm were dispersed. D‐zp37 Impedes Electron Transport Chain (ETC) and Triggers Abnormal Intracellular Reactive Oxygen Species (ROS) Accumulation We first monitored the change of zeta potential of VA 1579, VP 1647 and VV 5043 in the presence of D‐zp37 (Figure 4A). The upward trends from approximately −50 to −10 mV (VA 1579) or around −20 mV (VP 1647, VV 5043) with increasing concentration of D‐zp37 suggested that this peptide could interact with LPS to neutralize the membrane potential. To verify this hypothesis, LPS of VA 1579, VP 1647 and VV 5043 were first isolated and added to cultures containing fresh bacteria and 5 µm D‐zp37. We found that a higher concentration of exogenous LPS could effectively suppress the antibacterial activity of D‐zp37, resulting in better bacterial growth (Figure 4B). In addition, from the isothermal titration calorimetry (ITC) data, the calculated dissociation constant (Kd) between D‐zp37 and LPS was 4.8 ± 24.1 nm (Figure S1). (A) Zeta potential change of Vibrio cells before and after D‐zp37 treatment. (B) Growth kinetics of Vibrio cells in the presence of extra LPS. (C) OM permeability change of Vibrio cells. (D) IM permeability change of Vibrio cells. (E) Membrane potential change of Vibrio cells. (F) ATP level change of Vibrio cells. (G) ROS accumulation of Vibrio cells. (H) Interaction assessment between D‐zp37 and Vibrio DNA by gel retardation analysis. (I) Schematic illustration of the antibacterial mode of D‐zp37. The outer membrane (OM) and inner membrane (IM) permeability of G˗ve bacteria can be analyzed quantitatively by fluorescent dyes 1‐N‐phenylnaphthylamine (NPN) and SYTOX Green, respectively. From Figure 4C, a concentration of 2 µm D‐zp37 was sufficient to markedly destroy the integrity of the three studied Vibrio strains. Though the level of OM disruption was inferior to the positive control 0.1% Triton X‐100, D‐zp37 still showed strong membrane‐active capability. SYTOX Green spectrums further validated that the IM of Vibrio bacteria were significantly impaired after the D‐zp37 treatment (Figure 4D). Noteworthily, at least for VA 1579 and VP 1647, subjected to the treatment of 2 µm D‐zp37 could lead to more serious IM damage than 0.1% Triton X‐100. In addition to the physical structural damage of Vibrio bacteria, D‐zp37 was demonstrated to impede the ETC. Using a voltage sensitive probe 3, 3’‐dipropylthiadicarbocyanine iodide [DiSC3(5)], we noticed that the fluorescence intensity ​​increased rapidly within seconds in the presence of 1 µm D‐zp37 for all three Vibrio strains (Figure 4E). Compared with the acknowledged membrane potential modulator valinomycin (VAL), D‐zp37 could induce approximately the same degree of potential dissipation at 2 µm for VA 1579, VP 1647 and 8 µm for VV 5043. Furthermore, in comparation with untreated bacterial cells, the intracellular adenosine triphosphate (ATP) level of tested Vibrio strains decreased significantly after 2 µm D‐zp37 treatment (Figure 4F). Numerically, ATP contents of VA 1579, VP 1647 and VV 5043 all fell below 10% when the D‐zp37 concentration increased to 8 µm and beyond. Increased ROS levels were clearly detected when bacterial cells were exposed to D‐zp37 (Figure 4G). At their respective 0.25‐fold MICs, ROS levels were 1.42 times higher for VA 1579, 1.24 times higher for VP 1647 and 1.21 times higher for VV 5043, indicating that sub‐MIC treatment could already result in an obvious oxidative stress response. When Vibrio cells were treated with 4‐fold MIC of D‐zp37, the fold change of ROS increased sharply to 3.52 (VA 1579), 2.36 (VP 1647) and 3.35 (VV 5043), respectively. DNA of Vibrio strains were extracted and their interaction with D‐zp37 was examined. As illustrated in Figure 4H, incubating with 5 µm D‐zp37 led to slight tailing in agarose gel electrophoresis imaging. More significant DNA retardation was observed if higher concentrations of D‐zp37 were used, hinting that D‐zp37 was able to bind with DNA and cause aggregation. These results collectively depicted the multi‐mode antibacterial mechanisms of D‐zp37 against Vibrio (Figure 4I). Omics Studies Reveal that D‐zp37 Invalidates Membrane Stress Response and Blocks Amino Acid Biosynthesis of Vibrio Cells To further clarify the potential effects of D‐zp37 on signalling pathways of Vibrio cells, we employed transcriptomics and metabolomics approaches on VA 1579 before and after D‐zp37 treatment. Vibrio species, VA 1579, an environmentally isolated drug‐resistant strain showed the highest sensitivity to D‐zp37 was chosen for detailed analysis. Figure 5A shows the genomic circle map, serving as a reference for downstream molecular investigations. (A) Genomic circle map of VA 1579. Outermost circle: genome size marker. Second and third circles: genes on the plus strand and minus strand of the genome respectively. Different colours represent different clusters of orthologous genes functional classes. Fourth circle: repeat sequence. Fifth circle: tRNA is shown in blue and rRNA is shown in purple. Sixth circle: GC content. The region which GC content higher than the average GC content of the genome is shown in light yellow, the reverse is in blue. Innermost circle: GC‐skew, with dark gray representing regions where G content is greater than C and red representing regions where C content is greater than G. (B) Heat map of top 50 differential mRNA clustering. (C) Volcano plot of differential mRNA expression. Corresponding representative expressed proteins are annotated. (D) Partial least squares discriminant analysis between control and treat groups of non‐targeted metabolomics. (E) Volcano plot of differential metabolites. (F) Heat map of top 30 differential metabolites. Quinolinone (a): 2‐[4‐(3,4‐methylenedioxyphenyl) butyl]‐4(1H)‐quinolinone. Butanone (b): 4‐(2,6,6‐trimethyl‐1,3‐cyclohexadien‐1‐yl)‐2‐butanone. Indene (c): 5,7alpha‐dihydro‐1,4,4,7a‐tetramethyl‐4H‐indene. (G) KEGG enrichment scatterplot of non‐targeted metabolomics. Pathways with the top 20 smallest p‐values are displayed. (H) Network diagram of regulatory relationships between metabolites and pathways. Four key metabolites identified were shown in purple. Gene ontology and EggNOG database annotation of VA 1579 are presented in Figures S2 and S3, respectively. Using the complete genome of VA 1579 as a reference, a total of 4990 genes were identified, and Figure 5B lists the top 50 differential mRNA clustering between control and D‐zp37 treatment groups. Among them, 227 genes were upregulated, and 192 were downregulated (Figure 5C). Protein annotation of these genes highlights the potential involvement of the phage shock protein (PSP) family, a cell membrane stress protective system, in governing the antibacterial mechanisms of D‐zp37. Three members (pspG, pspB and pspA) were ranked as the top 5 downregulated genes, suggesting impaired membrane stress adaptation. In contrast, genes encoding ABC transporter permease were significantly upregulated in response to D‐zp37 exposure, indicating enhanced membrane transport activity. In validating the transcriptomics analysis, qPCR was employed to quantify the expression levels of the interested genes (Figure S4). The expression of pspG, pspB, pspA and DMT genes were all significantly downregulated, while the ABC transporter permease was significantly upregulated, agreeing with sequencing findings. However, the expression levels of SBPs were found to be downregulated, disagreeing with transcriptomics results. Next, metabolite changes were investigated. Classification annotation statistical charts of identified metabolites in the Human Metabolome Database and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were presented in Figures S5 and S6, respectively. Partial least squares discriminant analysis (PLS‐DA) suggested that 6 biological repeats in each group had a similar expression pattern (Figure 5D). The number of differential metabolic ions (Figure S7) and differential secondary metabolites (Figure S8) before and after D‐zp37 treatment was summarized. Volcano plot illustrated in Figure 5E demonstrated that among the secondary metabolites, 12 of them were upregulated and 11 were downregulated. The top 30 differential metabolites were displayed in a heat map (Figure 5F). Furthermore, KEGG enrichment revealed the possible biological functions of differential metabolites. According to Figure 5G, the amino acids biosynthesis of VA 1579, particularly valine, leucine and isoleucine, was significantly affected in the presence of D‐zp37. Based on KEGG pathway analysis of differential metabolites, a regulatory network diagram illustrating the relationships between key metabolites and associated pathways was constructed (Figure 5H). Four important metabolites, including pyruvic acid, L‐Valine, triethanolamine and L‐Histidinol, were highlighted due to their relevance in metabolic shifts. This network diagram effectively mapped the interactions between metabolic changes and pathway perturbation, highlighting the concerned pathways. Additionally, a correlation heat map revealing positive (shown in red) and negative (shown in blue) correlations between metabolites is presented in Figure S9, further supporting the systemic metabolic response induced by D‐zp37. D‐zp37 Reduces Bacterial Load of Vibrio Mixture In Vivo Vibrio can invade the human body via many routes, including from uncooked aquatic products, punctured wounds during seafood processing and seriously contaminated seawater [25], inducing gastritis, diarrhea and gangrene (Figure 6A). Here, we employed three infection models (shrimp, Galleria mellonella larvae and mouse) to comprehensively evaluate the antibacterial efficacy of D‐zp37. (A) Route of Vibrio infection and common vibriosis. (B) Live bacterial counts of infected shrimp samples during a 3‐day cold storage. (C) Virulence test of Vibrio mixture. (D) Survival ratio of infected larvae before and after D‐zp37 treatment. (E) Flow diagram of mouse model for in vivo antibacterial efficacy and histological studies (n = 6 per treatment group). (F) Body weight change of mouse. (G) Total Vibrio counts in the intestinal tract of the mouse. (H) H&E staining images of four groups of mice. Scale bar: 50 µm. (I) AB‐PAS staining images of four groups of mice. Scale bar: 50 µm. Shrimps were infected with Vibrio mixture to simulate food contamination and then treated with D‐zp37 and CAZ, respectively (Figure 6B). Time‐killing curves showed that when D‐zp37 was applied, total Vibrio colonies declined during a cold storage test lasting 3 days. The plummet occurred on the first day, which was consistent with previous bacteriolytic data. As controls, the CFU of PBS‐treated and CAZ‐treated Vibrio tended to increase. Next, Galleria mellonella larvae were used to preliminarily test the in vivo effect of D‐zp37. Three dosages (104–106 CFU per larva) of Vibrio mixture were injected to evaluate its virulence first (Figure 6C). After 12 h of infection, all individuals in the group infected with 105 CFU exhibited obvious melanization, suggesting that this dosage could lead to considerable virulence to larvae. Subsequently, the survival status of Galleria mellonella was monitored to see whether D‐zp37 could rescue infected larvae (Figure 6D). Inspiringly, the mortalities of D‐zp37‐treated groups were lower than those of the PBS‐treated group at each interval. After 48 h, the survival ratio of 50 µg/g group was >90%, far outperforming the PBS‐treated group (<20%). Vibrio infections bring a high risk to people, especially the immunocompromised ones. We therefore established an immunosuppressed mouse model to further verify the antibacterial effect of D‐zp37 in vivo (Figure 6E). Four mice groups were set and defined as: normal (healthy with normal feeding), immunosuppression (IMP, subjected to cyclophosphamide), bacterial infection (BI, subjected to cyclophosphamide, then Vibrio infection) and D‐zp37 (subjected to cyclophosphamide, then Vibrio infection, followed with 50 µg/g D‐zp37 treatment). Via intragastric (i.g) administration, mice were infected with high concentrations (OD600 = 0.3) of Vibrio mixture. Then, D‐zp37 was given by the same approach. As shown in Figure 6F, the average body weight of mice in the normal group increased, while mice in the other three groups suffered from immunosuppression lose weight obviously. Even though mice in the BI and D‐zp37 groups numerically had a greater weight loss compared with those in the IMP group, there was no statistical difference between these three groups. On the one hand, this indicated the successful establishment of an immunosuppressed model, and on the other hand, it suggested that the short‐term Vibrio infection (1 day) did not have much effect on the weight of mice. Differently, the plate counting result revealed that in the intestine of mice, live colonies of Vibrio cells varied significantly (Figure 6G). After the Vibrio infection, the bacterial number increased from an undetected level to >800 CFU/mg. The D‐zp37 treatment could reduce it to <200 CFU/mg. The capability of D‐zp37 in killing Vibrio in the complex gastrointestinal environment was thus confirmed. Besides the bacterial load, histology of the stomach, small intestine and large intestine were performed to investigate possible pathological changes after Vibrio infection. Hematoxylin and eosin (H&E) staining revealed that bacterial infection induced pathological changes of tissues in the stomach and intestine (Figure 6H). Alcian blue and Periodic acid‐Schiff (AB‐PAS) staining suggested that the bacterial infection may lead to excessive secretion of acidic mucous substance. After D‐zp37 treatment, some pathological features like organizational distortion were relieved (Figure 6I). D‐zp37 Ameliorates Vibrio Invasion‐Associated Gut Microbiota Disequilibrium Bacterial invasion may trigger gut microbiota disequilibrium [26]. We then analysed the fecal microbiomics of mice after Vibrio infection and D‐zp37 treatment (Figure 7A). Taking advantage of 16s rDNA‐based high‐throughput sequencing, microbial community composition in mouse fecal samples were studied. The Venn plots visually presented the number of amplicon sequence variants (ASVs) which were common or specific between each group (Figure 7B). The evolutionary tree illustrated in Figure 7C suggested that Firmicutes would be the dominant phylum, occupying 60% quota of the top 50 abundance species. Manhattan plots provided an overview of abundance change of different species between pairings (Figure 7D). For more details, the microflora of four groups in phylum level indicated that Vibrio infection could lead to ratio decrease of Firmicutes and increase of Bacteroidota, but D‐zp37 treatment reversed the trend (Figure 7E). In genus level, the abundance of several fundamental anaerobic acid‐producing bacteria like Odoribacter, Roseburia, and Oscillibacter declined after infection (Figure 7F). (A) Flow diagram of mouse model for gut microbiota analysis. (B) Venn plots of the number of ASVs. (C) Evolutionary tree of the top 50 abundance species. (D) Abundance changes of different species. (E) Microbiota variation between four groups in phylum level. (F) Microbiota variation between four groups in genus level. (G) Top 10 genera with a significant difference. (H) The correlation between pairwise dominant genera. (I) Phenotypes prediction including Gram staining category (a), biofilm forming (b) and potential pathogenicity (c), between four testing groups. By analysing the top 10 genera with significant difference, it should be noticed that Lachnospiraceae NK4A136 depleted dramatically in the BI group (Figure 7G(a)) but enriched again in D‐zp37 treatment group (Figure 7G(b)). There was no significant difference between the normal group and the D‐zp37 group (Figure 7G(c)). One of the key roles of Lachnospiraceae would be the efficient fermenters of complex carbohydrates and dietary fibers. They produce short chain fatty acids (SCFAs), notably acetic acid, propionic acid and butyric acid. Subsequent experiments validated that VA infection would cause a significant decrease in the content of these SCFAs in the feces of mice. However, after D‐zp37 treatment, their concentrations could be restored (Figure S10). Spearman network diagram reported the correlation between pairwise dominant genera (Figure 7H). Here we identified strong positive correlations between Oscillibacter and Colidextribacter, Clostridia and Firmicutes species. Negative correlations were also detected between Prevotellaceae and Eubacterium xylanophilum, Eisenbergiella and Enterorhabdus. Finally, the phenotypes of four groups were predicted by BugBase [27]. The three most concerned phenotypes, namely Gram staining category, biofilm forming and potential pathogenicity, were shown here. Ratio of G‒ve bacteria increased from 41% (normal) to 47% (IMP) and 59% (BI), respectively, but D‐zp37 treatment resulted in an obvious decrease to 42% (Figure 7I(a)). For the abundance of biofilm forming bacteria, these four groups didn't exhibit a significant difference (Figure 7(b)). However, the D‐zp37 treatment was predicted to lower the relative abundance of potential pathogenic bacteria (Figure 7I(c)). Discussion Although presenting desired helical structure and amphiphilic interface, the poor proteolytic stability of zp37 would certainly impose in vivo stability issues, limiting its potential clinical application. Non‐proteinogenic amino acids have unique advantages in resisting enzymatic degradation [28]. Thus, introducing unnatural residues is an efficient and widely used approach to prolong the in vivo half‐life of peptide drugs, maximizing their therapeutic effects [29]. Retaining the bioactivity of modified peptides is essential to ensure their therapeutic potential is not compromised. Based on the MIC data, D‐zp37 exhibited comprehensively enhanced bioactivity against Vibrio compared to its precursor, zp37. The susceptibility of tested Vibrio strains, which were resistant to penicillins and cephalosporins [30], to D‐zp37 suggested its potential as a promising antibacterial agent for overcoming antimicrobial resistance. HPLC results suggested that chiral conversion confers zp37 with improved stability. D‐zp37 also retained its bioactivity under various environmental conditions, with only a slight reduction in the presence of 50% FBS, 10% GJ and 25% IJ, demonstrating its strong environmental tolerance. Furthermore, D‐zp37 exhibited rapid bactericidal effects rather than bacteriostatic action, suggesting its potential for swift microbial eradication. Collectively, the improved proteolytic stability, broad‐spectrum bioactivity, and environmental resilience of the chirally‐converted peptide D‐zp37 highlighted its potential as a potent antibacterial candidate against Vibrio species. D‐zp37 exerts its bactericidal effect by disrupting the bacterial membrane structure. Encouragingly, this adverse impact is mild in mammalian cells. The selectivity is likely due to the composition differences in the membrane surface. While Vibrio membranes are rich in negatively charged phospholipids, mammalian cells predominantly contain neutral phospholipids [31]. After destroying the integrity of Vibrio’s membrane, D‐zp37 can be internalized into the bacterial cytoplasm, further enhancing its antibacterial efficiency via an intracellular bactericidal process. The ability of D‐zp37 to selectively lyse Vibrio cells without affecting the coexisting human digestive tract cells lays a strong foundation for its potential application in treating Vibrio‐induced gastrointestinal polymicrobial infections. The rapid motility of Vibrio cells, facilitated by flagella, contributes to the rapid dissemination of drug‐resistant genes and enhanced difficulty of prevention and control of the associated infections. The reduction of bacterial swimming area indicates that the presence of D‐zp37 can partially inhibit the spread of bacteria. This may be partly due to the direct bactericidal activity of D‐zp37, resulting in a decrease in the total number of bacteria participating in swimming behaviour. In addition, drug treatment may influence the physiological activities of bacteria, such as downregulating the expression of genes related to motility, impairing the flagellar function and movement [32]. Bacteria embedded within biofilms exhibit significantly greater resistance to antibacterial agents compared to their planktonic counterparts. In this study, we proposed that D‐zp37 may inhibit the efflux of polysaccharide from the interior to the exterior of Vibrio cells, leading to aberrant intracellular accumulation of polysaccharides. This disruption likely interferes with the extracellular matrix assembly, thereby delaying biofilm formation. Furthermore, D‐zp37 demonstrated the ability to disperse pre‐established biofilms, indicating its dual functionality in both prevention and eradication of biofilm‐associated Vibrio populations. The majority of cationic AMPs exert their antibacterial activity via targeting the bacterial membranes [33]. Their positively‐charged functional groups, such as amino or guanidine moieties, interact with negatively‐charged membrane components of G˗ve bacteria, particularly LPS, to disrupt essential physiological functions of bacteria [34]. In the case of D‐zp37, reduced antibacterial activity in the presence of exogenous LPS (Figure 4B) can be reasonably attributed to the sequestration of peptide molecules by free LPS, lowering the effective concentration of D‐zp37 available to act on the bacterial membrane and diminishing its killing efficiency. Membrane damage caused by AMPs is typically characterized by an increase in permeability [35]. Our results indicated that D‐zp37 compromises the proton motive force associated with ETC, thereby impeding energy production. This is supported by a marked decline in intracellular ATP levels (Figure 4F), indicating severe ETC dysfunction. Additionally, D‐zp37 induced excessive intracellular ROS accumulation, posing a critical threat to bacterial viability by oxidizing and irreversibly damaging essential cellular components [36]. Transcriptomics analysis further revealed that D‐zp37 disrupts the membrane stress response in Vibrio by impairing the PSP system, which is vital for maintaining cell viability under stress [37]. The downregulation of PSP‐related genes indicated compromised inner membrane integrity and a failure to initiate effective repair mechanisms. Concurrently, the significant upregulation of the gene encoding ABC transporter permease hinted enhanced membrane transport activity, possibly as a compensatory response to stress. This overexpression may also increase ATP consumption and imposed additional strain on membrane function. It is worth noting that qPCR data suggested the expression levels of SBPs were downregulated, contradicting the transcriptomics results. This discrepancy may reflect differences in temporal expression dynamics, post‑transcriptional regulation, or methodological sensitivity. Alternatively, D‑zp37 may trigger a rapid but transient activation of transporter permease genes while ultimately suppressing solute‑binding components as membrane damage progresses. Together, these findings suggest that D‑zp37 induces a multifaceted collapse of membrane homeostasis, disrupting both protective stress pathways and nutrient transport systems, contributing to reduced cellular resilience and promoting bacterial susceptibility. Complementary metabolomics data showed that D‐zp37 treatment led to significant metabolic shifts in Vibrio, particularly those involved in amino acid biosynthesis. The metabolic disruptions are likely to reflect as downstream consequences of the transcriptomic changes. Impaired membrane function and energy depletion (evidenced by reduced ATP levels) hindered biosynthetic processes, while stress‐induced transcriptional reprogramming may suppress genes involved in metabolic homeostasis, further contributing to the altered metabolite profile. The integrated analysis of molecular biological assays, transcriptomic profiles and metabolomic shifts present a consistent and multifaceted mechanism underlying the antibacterial activity of D‐zp37. Each dataset independently highlights disruptions in membrane integrity, energy metabolism, and stress response pathways. Together, they form a coherent picture of how the peptide compromises Vibrio viability. Structural damage observed through the biological assay is reflected in transcriptomic evidence of impaired membrane repair and metabolic shifts in biosynthetic functions, particularly amino acid metabolism. This alignment reinforces the reliability of the proposed mechanism: D‐zp37 initiates its lytic action by interacting with LPS on the outer membrane of Vibrio cells, compromising the membrane integrity and enabling entry into the cytoplasm. Once inside, the peptide disrupts the proton motive force and the ETC, leading to dissipation of membrane potential and reduced ATP synthesis. It also interacts with DNA, triggering abnormal intracellular accumulation of ROS, which oxidize and damage essential cellular components. The cumulative effects contribute to the rapid and effective killing of Vibrio cells. Importantly, this multifaceted mode of action may reduce the likelihood of resistance development as it targets multiple essential pathways simultaneously, making D‐zp37 an attractive antibacterial candidate. Building on the mechanistic insights, we next evaluated the therapeutic potential of D‐zp37 in living systems. The effectiveness of D‐zp37 in controlling foodborne contamination was first demonstrated using a shrimp model, confirming its applicability in real‐world food environments. Then, the in vivo curative effect of D‐zp37 was confirmed in a larvae infection model, where D‐zp37 treatment significantly improved the survival rates of larvae infected with a lethal dose of Vibrio bacteria. Encouraged by these results, a mouse model was employed to further assess the in vivo efficacy of D‐zp37. After intragastric administration, the number of live Vibrio bacteria in the intestine was significantly reduced, indicating that D‐zp37 retained its antibacterial activity even within the complex digestive tract environment. Fecal microbiome analysis revealed that Vibrio infection disrupted the synthesis of short‐chain fatty acids, primarily produced by anaerobic acid‐producing bacteria like Odoribacter. These metabolites play a vital role in maintaining intestinal health, and their reduction reflects a disturbance in gut microbial balance. For mice treated with D‐zp37, the abundance of Odoribacter and other SCFA‐producing strains was restored to baseline levels, suggesting that D‐zp37 contributes to the recovery of beneficial microbiota. Further supporting this observation, the abundance of Lachnospiraceae, which is a carbohydrate‐metabolizing bacterial family known for its role in promoting gut homeostasis [38], also returned to normal levels after D‐zp37 treatment. This indicates that D‐zp37 not only combats Vibrio infection but may also help preserve or restore the functional integrity of the gut ecosystem. Interestingly, no detectable Vibrio cells were found in the microbiome analysis, likely due to their low abundance relative to dominant endogenous gut bacteria. Collectively, these findings suggest that while Vibrio infection can compromise digestive health by disturbing microbial composition and metabolic output, D‐zp37 effectively mitigates these effects and supports the restoration of a healthy intestinal microbiota. Materials and Methods Peptide Characteristics and Antibacterial Activity Cleavage Sites Prediction Two peptides were customized from Synpeptide (Nanjing, China). The quality control result of D‐zp37 was shown in Figure S11 . Potential cleavage sites of zp37 were predicted by the website Expasy (https://www.expasy.org/resources/peptidecutter). Structure Prediction Secondary structure and Van der Waals surface diagram were generated by PEP‐FOLD 4 (https://bioserv.rpbs.univ‐paris‐diderot.fr/services/PEP‐FOLD4/). MIC Determination All Vibrio strains used in this study were isolated from the environment and stored in our lab. MIC values of zp37 and D‐zp37 against them were determined by a classic double‐dilution assay. In brief, bacterial suspensions (5 × 105 CFU/mL) in Mueller Hinton Broth (MHB, Hopebio, China) were treated with peptides at different concentrations for 18 h in a 96‐well plate at 37°C. The lowest concentration without observing bacterial turbidity was defined as its MIC. Proteolytic Stability in Trypsin Peptide solutions of zp37 and D‐zp37 (100 µm) were incubated with 10 µg/mL trypsin (HyClone, USA). Peak areas were monitored at 30 min intervals using HPLC (1200 infinity series, Agilent Technologies, USA). Peak area at 0 min was defined as 100%. Inhibition Effect in Different Conditions Selected Vibrio cells were inoculated in a 96‐well plate with an initial density of 5 × 105 CFU/mL in the absence or presence of specified additives. Trypsin and FBS were purchased from HyClone, USA. Simulated GJ and IJ were purchased from Chuang Feng Technology, China. NaCl was purchase from Sigma‐Aldrich, USA. After 6 h incubation at 37°C, OD600 values were measured (Clariostar, BMG Labtech, Germany). The turbidity value of bacteria incubated in the MHB medium was defined as 100% growth. Time Killing Test Selected Vibrio cells (approximately 106 CFU/mL) were treated with PBS (Solarbio, China) or D‐zp37 at their respective MIC, 2‐fold MIC and 4‐fold MIC. A volume of 10 µL bacterial suspension was sampled to perform plating counting in Mueller Hinton agar (Hopebio, China) at 20 min intervals, lasting 60 min. Bioactivity to Vibrio Mixture and Investigation on Cell Selectivity Turbidity Test OD600 values of three Vibrio strains (VA 1579, VP 1647 and VV 5043) were adjusted to approximately 1.0 individually in PBS. Then, they were mixed with the ratio of 1:1:1. Homogeneous polymicrobial suspension was divided into six groups and treated with PBS, 1–8 µm D‐zp37 or 8 µM CAZ (Sigma–Aldrich, USA) respectively. Turbidity changes of each group were monitored by measuring the OD600 values every 15 min by a biophotometer (D30, Eppendorf, Germany). SEM Homogeneous polymicrobial suspensions prepared as above were treated with PBS, 2 or 8 µM D‐zp37 respectively for 1 h at 37°C. Afterward, bacterial cells were fixed with 2.5% glutaraldehyde (Tokyo Chemical Industry, Japan) overnight at 4°C. Then, samples were immersed into a graded ethanol (Anaqua Chemicals Supply, USA) series from 50%, 70%, 90%, to 100% in sequence (10 min each grade). After resuspending them in 100% ethanol again, a volume of 2 µL suspension was dropped on a sterile cover glass. Finally, samples were air‐dried, coated with platinum, and imaged using a SEM facility (FEI Quanta 400F, Philips, Netherlands). Hemolysis Fresh blood was obtained from a healthy mouse. After centrifugation with 4000 rpm (Universal 320R benchtop centrifuge, Tegent, Hong Kong) for 10 min, the supernatant was discarded and the remaining red blood cells (RBCs) were diluted to 8% with PBS. RBCs were treated with PBS, D‐zp37 or 1% Triton X‐100 for 1 h, respectively. Cells were then centrifuged again at 4000 rpm for another 10 min. The absorbance of supernatant at 540 nm was measured (Clariostar, BMG Labtech, Germany). RBCs treated with PBS and 1% Triton X‐100 were set as negative control (0% hemolysis) and positive control (100% hemolysis) respectively. Triple Fluorescence Staining FITC‐D‐zp37 was customized from Synpeptide (Nanjing, China) with >95% purity (Figure S12). Homogeneous polymicrobial suspensions prepared as above was diluted to an OD600 value at 0.5. Bacterial cells were stained by 10 µg/mL DAPI (Thermo Fisher, USA) for 20 min, followed by 10 µM FITC‐D‐zp37 treatment. At specified timepoints (15, 30 and 60 min), part of cells was sampled, centrifuged to remove unbound dyes and resuspended in PBS containing 25 µm PI (Invitrogen, USA) for another 20 min staining. Subsequently, cells were rinsed thrice and fixed with 2.5% glutaraldehyde for 1 h. Prepared samples were dropped on a sterile coverslip and imaged by a live‐cell fluorescence imaging system (Eclipse Ti2E, Nikon, Japan). TEM Homogeneous polymicrobial suspensions prepared as above were treated with PBS and D‐zp37 at various concentrations for 1 h. Then, cells were subjected to dual fixation. The compositions of the prefixation solution included 2% formaldehyde (Sigma–Aldrich, USA), 2.5% glutaraldehyde, 0.075% ruthenium red (Sigma–Aldrich, USA) and 0.075 m lysine‐acetate (Sigma–Aldrich, USA) in cacodylate buffer (Sigma–Aldrich, USA) with pH at 6.9. After 20 min prefixation, bacterial cells were rinsed thrice with cacodylate buffer containing 0.075% ruthenium red and fixed again with the lysine‐acetate‐free mixture overnight at 4°C. On the next day, secondary fixation was performed using the solution composed of 1% OsO4 (Sigma–Aldrich, USA) and 0.075% ruthenium red in cacodylate buffer for another 1 h. Finally, the samples were dehydrated with a series of graded ethanol solutions. After resin embedding, and ultrathin sectioning, TEM (HT7700, Hitachi, Japan) pictures were taken with 12000‐ and 30000‐fold magnification. Cytotoxicity GES‐1 cells seeded in a 96‐well plate were incubated with D‐zp37 at various concentrations for 24 h, followed by adding 10 µL of CCK‐8 solution (Solarbio, China) for another 4 h incubation. Next, the absorbance values at a wavelength of 450 nm were measured using a microplate reader (Clariostar, BMG Labtech, Germany). GES‐1 cells treated with PBS was defined as 100% survival. Co‐Culture Image A sterile coverslip was placed at the bottom of a 6‐well plate. GES‐1 cells in Dulbecco's modified eagle medium (DMEM) containing 10% FBS were seeded. After 24 h incubation, Nile red (Solarbio, China) at 10 µg/mL was then employed to stain GES‐1 cells for 30 min. Subsequently, DMEM was replaced by a homogeneous polymicrobial suspension with an OD600 value of 0.3 for 10 min infection. Next, the bacterial solution was gently removed and such a co‐culture system of Vibrio and GES‐1 was treated with 10 µg/mL Hoechst 33258 (Solarbio, China) for 30 min and 20 µM FITC‐D‐zp37 for another 30 min in succession. Afterward, Vibrio cells were collected and rinsed thrice. In the meanwhile, GES‐1 cells were also rinsed thrice. The coverslip, whose upper surface was covered with GES‐1 cells, was taken out and a total of 2 µL washed Vibrio solution was dropped on top of that. Finally, the photos were filmed by a live‐cell fluorescence imaging system (Eclipse Ti2E, Nikon, Japan). Motility and Biofilm Related Studies Swimming The 0.5% Luria–Bertani (LB) agar (Hopebio, China) plates containing different concentrations of D‐zp37 (0–32 µm) were prepared in advance. A volume of 10 µL homogeneous polymicrobial suspension with OD600 at 0.1 was dropped on the center gently. Swimming pictures from day 1 to day 6 were captured by the ChemiDoc Imager (Bio–Rad, Hercules, USA). The longest straight‐line distance through the center was defined as the swimming diameter. Exopolysaccharide (EPS) Content Analysis Homogeneous polymicrobial suspension (OD600 = 0.2) was divided into four groups and seeded in a black 96‐well plate. Vibrio mixtures were then treated with PBS and 2, 8 or 16 µm D‐zp37 for 24 h, respectively. Planktonic cells were removed and wells were rinsed thrice. Next, the sediment was stained with 20 µg/mL concanavalin A conjugated by Alexa Fluor 488 (Thermo Fisher, USA) for 30 min to label EPS. Qualitative images were filmed by a live‐cell fluorescence imaging system (Eclipse Ti2E, Nikon, Japan) and the quantitative analysis was performed by a microplate reader (Clariostar, BMG Labtech, Germany). Intracellular Polysaccharide Content Analysis Similar with the protocol described above, homogeneous polymicrobial suspension (OD600 = 0.2) was treated with PBS and 1, 2, 4 or 8 µm D‐zp37 for 4 h, respectively. Cells were then centrifuged and bacterial pellets were collected. They were resuspended in equivalent PBS containing 20 µg/mL concanavalin A conjugated by Alexa Fluor 488 for 30 min staining. After that, OD600 values and fluorescence units were measured (Clariostar, BMG Labtech, Germany). Biofilm Inhibition Homogeneous polymicrobial suspensions (initial total CFU was 5 × 105 per mL) were treated with different concentrations of D‐zp37 for 18 h in a 96‐well plate (150 µL per well). Planktonic cells were gently removed, and all wells were washed three times by deionized water. Then, biofilms were stained with 0.1% crystal violet (CV) solution (Sigma‐Aldrich, USA) for 30 min. After rinsing, a volume of 150 µL of absolute ethanol was added to dissolve the CV bound to biofilms. The control group without D‐zp37 treatment was defined as 100%. The relative amount of biofilm for each group was calculated by comparing the absorbance at 595 nm with the control group. Five biological repeats were performed in this study. Biofilm Eradication Homogeneous polymicrobial suspensions (initial total CFU was 5 × 105 per mL) were seeded in a 96‐well plate and incubated at 37°C for 24 h. After removing planktonic bacterial cells, sediments were rinsed gently followed by 6 h D‐zp37 treatment (0–64 µm). Phase images were subsequently filmed (Eclipse Ti2E, Nikon, Japan). Antibacterial Mechanism Zeta Potential Vibrio cells at logarithmic phase were collected and resuspended in PBS to OD600 at 0.2. Then, cells were subjected to D‐zp37 (0‒8 µm) treatment for 10 min. The zeta potential of each sample was measured by Delsa Nano C particle analyser (Beckman Coulter, USA). Growth Kinetics in the Presence of LPS LPS of VA 1579, VP 1647 and VV 5043, was extracted following the instructions of the specification of LPS extraction kit (Bestbio, China). Subsequently, bacterial suspensions of three Vibrio cells (OD600 = 0.1) were divided into four groups and treated with PBS, 10, 20 and 40 µg/mL corresponding LPS, respectively. The OD600 values were recorded continuously for 4 h (Clariostar, BMG Labtech, Germany). ITC D‐zp37 (200 µg/mL) and LPS (20 µg/mL) were dissolved in water, respectively. D‐zp37 solution was then titrated into LPS solution following the handbook of Malvern MicroCal PEAQ‐ITC Automated Ultrasensitive Isothermal Titration Calorimeter (Malvern Instruments, UK). The operation parameters were set as –19 injections, 0.5 µL (first) and 2 µL (subsequent) for each injection volume, stir rate at 750 rpm, and temperature at 25°C. NPN Assay Vibrio cells (OD600 = 0.6) were treated with D‐zp37 at various concentrations (0.5‒32 µm) for 1 h. Cells treated with PBS and 0.1% Triton X‐100 (Solarbio, China) were set as two control groups. Then, bacterial suspensions were stained by 10 µm NPN (Sigma–Aldrich, USA) for 5 min and the fluorescence units were tested with an exciting wavelength at 350 nm and an emission wavelength at 420 nm (Clariostar, BMG Labtech, Germany). SYTOX Green assay Membrane permeability was monitored following the instructions of the specification provided by the vendor (Mao Kang biotechnology, China) with slight modification. Briefly, Vibrio cells (OD600 = 0.6) were treated with PBS, D‐zp37 or 0.1% Triton X‐100 respectively for 1 h. Next, SYTOX Green (5 µm) was applied to stain for 10 min at room temperature. Emission spectrums from 500 to 600 nm were recorded under the excitation wavelength at 488 nm (Clariostar, BMG LABTECH, Germany). Membrane Potential Vibrio cells at the logarithmic phase were collected and adjusted to OD600 value of 0.3 in PBS. Then, cells were incubated with 0.5 µm potentiometric probe DiSC3(5) (Thermo Fisher, USA) for 1 h. Next, they were subjected to the treatment of PBS, D‐zp37 (1–32 µm) and VAL (10 µg/mL), respectively. After that, the change of fluorescence units was instantly recorded with an excitation wavelength of 610 nm and an emission wavelength of 660 nm (Clariostar, BMG Labtech, Germany). ATP Level Vibrio cells at the logarithmic phase were collected and adjusted to OD600 value of 0.5 in PBS. Cells were then treated with PBS or D‐zp37 at various concentrations for 1 h. After that, cells were lysed and the relative ATP levels were determined by ATP assay kit (S0026, Beyotime, China) following the instructions of the specification. The luminescence value emitted by PBS‐treated group was defined as 100%. ROS Accumulation Vibrio cells were resuspended in PBS with an adjusted OD600 value of 0.6. The cells were divided into six groups and treated with 0.25‐fold, 0.5‐fold, 1‐fold, 2‐fold, 4‐fold MIC of D‐zp37, and PBS as the control for 1 h respectively. Next, cells were stained with 10 µM 2, 7‐dichlorodi‐hydrofluorescein diacetate (DCFH‐DA, Thermo Fisher, USA) for 30 min. After thrice rinse, fluorescence units were recorded with an exciting wavelength of 488 nm and an emission wavelength of 525 nm (Clariostar, BMG Labtech, Germany). DNA Binding DNA of three Vibrio strains was obtained by DNA extraction and purification kits (Thermo Fisher, USA). They were then mixed with an equal volume of D‐zp37 solutions and the final concentrations of D‐zp37 were set as 0, 5, 10, 20, 40 and 80 µm respectively. After 1 h incubation, DNA shift was measured by 0.8% agarose gel electrophoresis. Transcriptomics and Metabolomics Whole‐Genome Sequencing VA 1579 was isolated and purified. Cells at logarithmic phase were sent to Shenzhen SMQ group medical laboratory (Shenzhen, China) for sequencing using PacBio technology. Transcriptomics and Real‐Time PCR Analysis VA 1579 cells were grown to the logarithmic phase (OD600 = 0.6) and treated with PBS or D‐zp37 (1 µm) for 2 h. Cells were then collected and sent to Shenzhen SMQ group medical laboratory (Shenzhen, China) for transcriptomics analysis. The expression level of genes identified to be significantly altered were quantified with Real‐Time qPCR. Total RNA was extracted from treated cells using the StarSpin Fast Cell RNA Kit (Genstar), following the manufacturer's instructions. Removal of genomic DNA contamination and cDNA synthesis were performed simultaneously with StarScript Pro All‐in‐one RT Mix with gDNA Remover Kit (Genstar, Beijing, China). The synthesized cDNAs was then quantified with RealStar Fast Pro SYBR qPCR Mix Kit (Genstar, Beijing, China) on Roche LightCycler 480 II Real‐Time PCR System equipment (F. Hoffmann‐La Roche Ltd, Basel, Switzerland) with primers detailed in Table S1. Data were normalized with the endogenous RpoD gene [39], which encode RNA polymerase sigma factor. The relative gene expression levels were calculated using the 2–ΔΔCt relative quantification method [40]. All experiments were repeated in triplicate. Metabolomics VA 1579 cells at logarithmic phase (OD600 = 0.6) were divided into two groups (6 biological repeats per group). One was treated with 1 µm D‐zp37 for 1 h, another was treated with equivalent PBS. Cells were then centrifuged to collect and washed with PBS thrice. After that, Vibrio cells were resuspended in −80°C methanol and stored at −20°C overnight. Finally, cells were centrifuged at 12000 rpm for 15 min. Supernatant was collected and sent to LC Biotechnology CO., Ltd (Hangzhou, China) for metabolomics analysis. In Vivo Antibacterial Efficacy Shrimp Model Live shrimps (Trachypenaeus curvirostris) were purchased from Tai Po market in Shatin. They were immersed in the homogeneous polymicrobial suspension (OD600 = 0.6) for 30 min for Vibrio infection. Shrimps were then divided into four groups and D‐zp37 or CAZ (40 µg/g) were added. Samples were mixed evenly and stored at 4°C. At indicated intervals, an aliquot of shrimps (2 g) was transferred into 10 mL of sterile PBS and subjected to adequately grind. CFUs of total Vibrio cells in prepared homogenate was calculated by plate counting assay in the selective thiosulfate‐citrate‐bile salts‐sucrose (TCBS) agar (Hopebio, China). Vibrio Virulence Galleria mellonella larvae (0.3‒0.4 g) were randomly divided into four groups. A volume of 10 µL PBS or homogeneous polymicrobial suspensions (104, 105 and 106 CFU) were injected into larvae via the final prolegs. After 12 h incubation at 37°C, the survival ratio was recorded. Antibacterial Efficacy in Galleria Mellonella Model Except for the control group, Galleria mellonella larvae were injected with 105 CFU of homogeneous polymicrobial suspension. After 1 h infection, larvae were treated with PBS or D‐zp37. Survival ratios were determined at the specific time‐point. Each group had 12 larvae. Antibacterial Efficacy and Histological Analysis in Mouse Model The mouse experiments were approved by the animal ethics committee, the Hong Kong Polytechnic University (approval number 23‐241645‐FSN‐R‐GRF). Female BALB/c mice (18–20 g) were divided into four experimental groups (n = 6): (1) Normal group, consisting of healthy mice; (2) IMP group, with mice subjected to immunosuppression without Vibrio infection; (3) BI group, with mice undergoing immunosuppression followed by Vibrio infection; (4) D‐zp37 group with mice undergoing immunosuppression and Vibrio infection, followed by treatment with 50 µg/g D‐zp37. Immunosuppression was achieved by injecting 250 µg/g cyclophosphamide (Meryer, China) intraperitoneally (IMP, BI and D‐zp37 groups). Three days later, a volume of 100 µL homogeneous polymicrobial suspension (OD600 = 0.3) was given to mice intragastrically (BI and D‐zp37 groups). After 1 h infection, 50 µg/g D‐zp37 was applied to treat D‐zp37 group of infected mice intragastrically. For mice from other groups, they were given equivalent volume of PBS. Body weight of all mice was constantly monitored from the day of performing immunosuppression to one day after D‐zp37 treatment. Then, mice were sacrificed, and the large intestine was separated, subsequently homogenized. Total CFU of Vibrio were counted using TCBS agar. In the meanwhile, gastrointestinal tissues including stomach, small intestine and large intestine were fixed in 4% paraformaldehyde (Sigma–Aldrich, USA) and sent to Servicebio Technology (Wuhan, China) for histological analysis based on H&E and AB‐PAS staining. Gut Microbiota Fecal Sample Collection Mice were divided into four experimental groups (n = 3): (1) Normal group, consisting of healthy mice; (2) IMP group, with mice subjected to immunosuppression without Vibrio infection; (3) BI group, with mice undergoing immunosuppression followed by Vibrio infection; (4) D‐zp37 group with mice undergoing immunosuppression and Vibrio infection, followed by treatment with 50 µg/g D‐zp37. For BI and D‐zp37 groups, a homogeneous polymicrobial suspension was adjusted to OD600 value at 0.2. Then the bacterial mixture was diluted by 20‐fold in PBS. Mice were then infected with a volume of 100 µL of the diluted bacterial mixture intragastrically. For D‐zp37 group, another 50 µg/g D‐zp37 was given after 30 min infection. 24 h later, when mice defecated, fresh fecal samples were collected from the rectal area directly and cryopreserved at −80°C. Gut Microbiota Gut microbiota was analysed by Biotree, Shanghai using 16S rDNA sequencing technology. DNA from four fecal samples was extracted, followed by PCR amplification. Purified samples were finally sequenced on the NovaSeq platform (Illumina, USA) according to the manufacturer's recommendations. SCFA Concentration The fecal samples from four mice groups were collected. The acetic acid, propionic acid and butyric acid with known concentrations were used to draw the standard curves. Then, a volume of 2 mL of water (1:3 phosphoric acid solution) was added to each fecal samples, followed by vortex and homogenize for 2 min. Subsequently, 1 mL of ether was added for extraction for 10 min. Samples were centrifuged (4000 rpm, 20 min, 4°C). Another 1 mL of ether was supplemented for an additional 10 min centrifugation. Liquids from two extraction steps were collected, mixed and evaporated to less than 1 mL for analysis by gas chromatography‐mass spectrometer (GC‐MS, Thermo Scientific ISQ LT, USA). Statistical Analysis The results were expressed as mean ± standard error. Unless specified, three biological repeats were performed in all experiments. Original data were analysed by ANOVA and t‐test, where ns suggested no significant difference, * p < 0.05, ** p < 0.01, *** p < 0.001, and **** p < 0.0001. Conclusion In this work, a new therapeutic peptide, namely D‐zp37, against polymicrobial infection caused by Vibrio superbugs was proposed. This novel peptide employed a simple chirality conversion strategy to improve its antibacterial activity and proteolytic stability. It could lyse Vibrio cells rapidly without affecting co‐existed normal cells. Mechanistic studies revealed that D‐zp37 impeded ETC and could restrain the biofilm establishment of Vibrio cells via inhibiting the efflux of intracellular polysaccharides. Omics studies further suggested that D‐zp37 may invalidate membrane stress response and blocks amino acid biosynthesis of Vibrio cells. In addition, the excellent antibacterial efficacy of D‐zp37 was verified in three different infection models, including a shrimp contamination, a Galleria mellonella larvae infection and an immunosuppressed mice intestinal infection model. The intragastric administration of D‐zp37 was further found to restore the imbalanced gut microbiota of immunosuppressed mice induced by polymicrobial Vibrio invasion. Collectively, these findings could be beneficial to people who are suffering from refractory Vibrio polymicrobial infections, especially those coastal residents with compromised immune systems. Author Contributions Ping Zeng: Conceptualization, Methodology, Formal analysis, Investigation, Visualization, Writing – original draft. Qipeng Cheng: Methodology, Formal analysis, Investigation, Funding acquisition. Xiaoxu Zhang: Methodology, Investigation. Honglan Wang: Methodology, Investigation. Pengfei Zhang: Methodology, Investigation. Jinhan Zhang: Methodology, Investigation. Xinyi Ding: Validation, Investigation. Lanhua Yi: Methodology. Kwok‐Yin Wong: Resources. Kin‐Fai Chan: Resources. Sheng Chen: Resources. Sharon Shui Yee Leung: Resources, Writing – review and editing, Supervision, Project administration, Funding acquisition. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Supporting File: advs75335‐sup‐0001‐SuppMat.docx Acknowledgements The authors are thankful to Dr. Junju Zou and Dr. Qianbo Song for their kindly help regarding microbiota analysis. The authors are thankful to Ms. Xuejiao Song for helping with the ITC test. The authors would like to acknowledge the financial support by the Health and Medical Research Fund Hong Kong (Grant Number 23220102), the National Natural Science Foundation of China (Grant Number 32202987), and the Key Program of Anhui Educational Committee (Grant Number 2022AH050179). Zeng P., Cheng Q., Zhang X., et al. “A Chirality‐Converted Bacteriolytic Dodecapeptide Regulates Vibrio‐Induced Polymicrobial Infection and Ameliorates Invasion‐Associated Gut Microbiota Disequilibrium.” Advanced Science 13, no. 38 (2026): e22326. 10.1002/advs.202522326 Ping Zeng and Qipeng Cheng are co‐first authors. Contributor Information Sheng Chen, Email: sheng.chen@polyu.edu.hk. Sharon Shui Yee Leung, Email: sharon.leung@cuhk.edu.hk. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. References Associated Data Supplementary Materials Supporting File: advs75335‐sup‐0001‐SuppMat.docx Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.