A comparative analysis of publicly available oyster plasma peptide raw data was first conducted to reconstruct the oyster plasma peptide dataset. To re-screen peptides with antimicrobial activity, multiple online antimicrobial peptide activity prediction models/databases were employed here to conduct a large-scale joint analysis of antimicrobial activity on this peptide dataset. Specifically, we first extracted potential antimicrobial peptides (AMPs) predicted by dbAMP (https://awi.cuhk.edu.cn/~AMPActiPred/). Subsequently, we applied these peptides to the CAMP (Random Forest, RF) (https://camp.bicnirrh.res.in/) and Antimicrobial Peptide Scanner vr.2 (https://www.dveltri.com/ascan/) for scoring. Peptides that scored above 0.5 in both databases were selected as the new dataset. These selected peptides were chemically synthesized by Sangon Biotech (Shanghai) Co., Ltd. to produce peptide reagents for antimicrobial activity experiment.

In conjunction within silico tool assessments (htxxps://aps.unmc.edu/design/improve) aimed at improving antimicrobial effectiveness, we developed a new peptide, named OAMP, by optimizing the amino acid composition of peptide AMP12. The peptide was then evaluated against entries in the APD3 databases. Sequence alignments of OAMP with AMP42 and Arenicin-1 (AP04610) from the marine polychaeta Arenicola marina [24], were performed using BioEdit software. The tertiary structure of OAMP was analyzed using the online peptide structure prediction software PEP-FOLD (https://bioserv.rpbs.univ-paris-diderot.fr/services/PEP-FOLD3/). Subsequently, hydrophobic structure analyses were conducted using HeliQuest (https://heliquest.ipmc.cnrs.fr/), focusing on hydrophobicity, hydrophobic moment, and net charge (z).

To evaluate the antimicrobial activities of the peptides, approximately 10^6 to 10^7 CFU/mL of bacterial suspension was prepared. Peptide solution in final concentration of 0, 2.5, 5, 10, 25, 50 μM was co-incubated with the bacterial suspension for 1 hour, respectively. Subsequently, an equal volume of the peptide-bacterial suspension was spread onto agar plates, followed by overnight incubation for colony formation. The control group received an equal volume of PBS instead of the peptide solution. The percentage survival rate compared to the control group was determined. Each experiment was repeated three times, with three replicates in each repetition.

Minimum inhibitory concentrations (MICs) of OAMP and antibiotics were determined using broth microdilution method. In brief, under aseptic conditions, different concentrations of antimicrobial agents were serially diluted and added to sterile 96-well polystyrene plates. 10 μl of each drug solution was added from the 1st to the 11th well. Subsequently, a suspension of CR-AB bacteria prepared either by the broth dilution or direct inoculum method, adjusted to a concentration equivalent to 0.5 McFarland turbidity standard, was diluted 1:1000 in MH broth and 100 μl was added to each well. The plates were sealed and incubated at 35°C under ambient air for 16 to 20 hours before interpreting the results. The concentrations of OAMP tested were 64, 32, 16, 8, 4, 2, 1, and 0 μM. The concentrations of antibiotics tested were 32, 16, 8, 4, 2, 1, 0.5, 0.25, 0.13, 0.06, 0.03, and 0 mg/L.

Referring to the combination therapy of OAMP and antibiotics, a checkerboard assay was conducted to determine the Fractional Inhibitory Concentration Index (FICI) against five clinical CR-AB strains, using a 96-well plate setup based on established methods. In summary, 2-fold serial dilutions of OAMP were arranged in Columns 1 to 11, while rows A to G contained 2-fold serial dilutions of antibiotics (meropenem, tigecycline, and colistin, respectively). Column 12 contained a serial dilution of antibiotics alone, and row H contained a serial dilution of OAMP alone. The concentrations of OAMP tested were 0, 1, 2, 4, 8, 16, 32, and 64 μM, and the concentrations of antibiotics ranged from 32 to 0 mg/L (32, 16, 8, 4, 2, 1, 0.5, 0.25, 0.13, 0.06, 0.03, and 0 mg/L). Fifty microliters of each concentration were added to the wells of a sterile microplate, followed by the addition of 100 μl of bacterial suspension (5×10^7 CFU/mL). After overnight incubation, growth was assessed, and the minimum inhibitory concentration (MIC) was determined as the lowest concentration at which no bacterial growth occurred.

Interpretive criteria follow CLSI 2023 (meropenem MIC ≥4 mg/L for resistance), FDA (tigecycline MIC ≤0.5 mg/L for susceptibility), and EUCAST 2023 (colistin MIC ≤2 mg/L for susceptibility). The interaction between OAMP and antibiotics was evaluated using FICI, which categorizes their combined effects into four types based on the calculated FICI value:

A and B are the MIC of each drug in combination (in a single well), and MIC_A and MIC_B are the MIC of each drug individually. Synergy: FIC ≤ 0.5, indicating that the combined activity of the antibiotics is significantly greater than the sum of their individual activities. Additivity: FIC > 0.5 to ≤ 1, where the combined activity equals the sum of their individual activities. Indifference: FIC > 1 to ≤ 4, showing that the combined activity equals the activity of the individual antibiotics. Antagonism: FIC > 4, indicating that the activity of one antibiotic is significantly reduced by the presence of the other.

In vitro hemolytic activity of the antimicrobial peptide OAMP was evaluated using a standardized protocol. Sheep red blood cells and chicken red blood cells (RBCs) were collected and washed with phosphate-buffered saline (PBS) to remove plasma and buffy coat, respectively. RBCs were then suspended in PBS to achieve a 2% (v/v) concentration. Serial dilutions of OAMP were prepared in PBS. In a sterile 96-well U-bottom microplate, 50 μL of each OAMP dilution was added to designated wells as experimental groups. Negative control wells contained 50 μL of PBS, and positive control wells contained 10% Triton X-100 to induce complete hemolysis. Each well received 1 × 10^5 RBCs suspension (50 μL total volume). The microplate was then incubated at 37°C with 5% CO₂ for 30 minutes. After incubation, the microplate was centrifuged and 100 μL of supernatant from each well was transferred to a new flat-bottom 96-well microplate. Absorbance was measured at 450 nm using a microplate reader (Bio-Rad) to determine the optical density (OD). The percentage of hemolysis was calculated using the formula: % Hemolysis = [(OD450_sample - OD450_negative control)/(OD450_positive control - OD450_negative control)] × 100. The experiment was performed in triplicate, and results were reported as mean ± standard deviation (SD).

The cytotoxicity of the antimicrobial peptide (OAMP) was evaluated using RAW264.7 murine macrophage leukemia cells. Briefly, RAW264.7 cells were seeded into 96-well cell culture plates at a density of 1 × 10^5 cells per well and allowed to adhere overnight. After discarding the supernatant, cells were washed twice with PBS. Serial dilutions of OAMP ranging from 0 to 32 μM were prepared in cell culture media. Subsequently, 100 μL of each OAMP dilution was added to the respective wells containing the washed RAW264.7 cells. Each condition was performed in triplicate wells. Following a 24-hour incubation period under 37°C and 5% CO₂, 10 μL of CCK-8 solution was added to each well and further incubated for 2 hours. Absorbance was measured at 450 and 650 nm using a microplate reader (Bio-Rad) to determine the optical density (OD). Cell viability was calculated based on the absorbance values relative to the negative control wells: cell viability % = [(A_sample - A_650)/(A_negative control - A_650)] × 100), where A_sample is the absorbance (at 450 nm) of the treated cells, A_650 is the background absorbance (at 650 nm), and A_negative control is the absorbance of the negative control. Results were expressed as mean ± standard deviation (SD).

The marine bivalve-derived peptide library constitutes a valuable natural repository of bioactive peptides (Mao et al., 2021). Capitalizing on recent advances in antimicrobial peptide research and the development of novel computational prediction methodologies, we systematically reanalyzed the published oyster hemolymph peptide dataset (PRIDE partner repository: PXD025247) (Mao et al., 2021) using state-of-the-art online prediction platforms, AMPAPred, CAMP, and AMP scanner vr2. Through comprehensive screening based on antimicrobial activity scoring metrics (Figure 1), 36 high-scoring peptide candidates were selected for development as next-generation antimicrobial agents (Table 1).

The antimicrobial activity screening of the 36 candidate peptides against four pathogens - Escherichia coli (ATCC 25922), Pseudomonas aeruginosa (ATCC 27853), Streptococcus pneumoniae (ATCC 49619), and Staphylococcus aureus (ATCC 29213) - revealed three peptides (AMP07, AMP12, and AMP29) exhibiting varying degrees of inhibitory effects (Table 2). Of these, AMP12 demonstrated particularly potent and consistent activity across both Gram-negative and Gram-positive strains, with minimum inhibitory concentrations (MICs) of 10 μM against P. aeruginosa, 5 μM against E. coli, 5 μM against S. pneumoniae, and 25 μM against S. aureus. The superior efficacy profile of AMP12, established it as the most promising lead compound for further investigation.

Bioinformatic analysis using the Antimicrobial Peptide Database (APD3; https://aps.unmc.edu/design/input) revealed that AMP12 exhibits 50% sequence homology with Arenicin-1 (AP04610) (Orlov et al., 2019), a potent antimicrobial peptide derived from the marine polychaete Arenicola marina (Figure 2A). However, comparative evaluation demonstrated that AMP12’s antimicrobial activity (MIC range: 5-25 μM) was significantly weaker than that of Arenicin-1 (MIC: 1-4 μM) against comparable bacterial strains. Therefore, based on established structure-activity relationships demonstrating that strategic modifications of charge distribution and hydrophobicity can enhance AMP efficacy (Gagat et al., 2024), we engineered an optimized derivative through two key mutations: (1) deletion of neutral asparagine at position 11 (ΔN11) to increase overall hydrophobicity, and (2) substitution of threonine with lysine at position 17 (T17K) to augment cationic charge density. This rational design yielded the optimized peptide OAMP (KVAFVKVVKVLGCFKKR), with predicted improvements in both GRAVY (Grand Average of Hydropathy) and net charge (Figure 2B). According to APD3, OAMP exhibits a GRAVY of 0.61 and possesses a total net charge of +6.

OAMP exhibits optimized structural characteristics that synergistically enhance its antimicrobial efficacy, as demonstrated through comprehensive physicochemical characterization. The peptide’s sequence exhibits a strategically optimized composition with: (i) seven hydrophobic residues (Val×5, Leu×1, Phe×2, Ala×1) that enhance membrane penetration, (ii) six positively charged lysine residues (Lys×5) and one arginine (Arg×1) for strong electrostatic attraction to microbial membranes, while (iii) completely lacking negatively charged residues to prevent charge neutralization. Computational modeling using PEP-FOLD 3.5 confirmed the formation of stable α-helical structures (Figure 2C), a conformation known to facilitate membrane disruption. Subsequent analysis of its physicochemical properties and amino acid distribution via HELIQUEST illustrated a hydrophobic face on the helical wheel diagram (Figure 2D). This carefully balanced architecture between hydrophobic and cationic domains, coupled with its stable secondary structure, positions OAMP as a promising antimicrobial candidate with optimized membrane-targeting capabilities.

Scanning electron microscopy revealed specific bactericidal mechanisms of OAMP against Gram-positive and Gram-negative pathogens (Figure 3). In S. aureus, OAMP treatment induced the formation of nanopores characteristic of the barrel-stave model, while maintaining overall cellular integrity - suggesting selective membrane disruption without complete lysis. Conversely, V. parahaemolyticus exhibited multifocal outer membrane collapse and cytoplasmic shrinkage in cells, indicative of LPS layer penetration followed by content leakage. Control groups of both species maintained smooth, intact morphologies, confirming OAMP’s specific membrane-disrupting activity.

To evaluate the therapeutic potential of OAMP against drug-resistant pathogens, we first assessed the antibacterial activity of OAMP against five clinical isolates of carbapenem-resistant Acinetobacter baumannii (CR-AB) strains collected from the Shenzhen Luohu People’s Hospital over the past two years. Using broth microdilution method, we first determined the minimum inhibitory concentrations (MICs) of three commonly used antibiotics (meropenem, tigecycline, and colistin) against these strains (Table 3, MIC_B). Our findings indicate that all five CR-AB strains exhibited resistance to meropenem, while they showed susceptibility to tigecycline and colistin. Subsequently, we evaluated the antibacterial activity of OAMP against the five drug-resistant strains. In three independent experimental replicates, OAMP demonstrated MIC values of 4 μM or 8 μM against strains 1 and 2, while strains 3, 4, and 5 exhibited an MIC of 4 μM (Table 3, MIC_A). These results indicate that OAMP exhibits varying degrees of ability against different CR-AB strains, suggesting its potential utility in combating these pathogens.

To determine whether OAMP has the potential for combination therapy with antibiotics, according to previously reported assay (Chevalier et al., 1995; Garcia, 2010), we employed the broth microdilution checkerboard method to assess the interactions of three antimicrobial combinations: meropenem + OAMP, colistin + OAMP, and tigecycline + OAMP. Bacterial growth status was determined by visual turbidity readings, as shown in Figure 4. The MIC values for combination therapy (OAMP/antibiotic) were systematically tabulated in Table 3. Compared to individual drug MICs (MIC_A for OAMP and MIC_B for antibiotics), all combination pairs demonstrated reduced MIC values (A/B), indicating synergistic effects that allow dose reduction of both agents while maintaining therapeutic efficacy - a strategy that may mitigate antibiotic-associated adverse effects.

Then, the effectiveness of antimicrobial combinations was assessed based on the Fractional Inhibitory Concentration Index (FICI, Table 4). According to the FICI interpretation criteria (Odds, 2003), the interactions between OAMP and different antibiotics were primarily additive (FICI = 0.5~1), with only synergy observed with meropenem in the treatment of CR-AB_2 strain (FICI = 0.3125).

To further analyze the potential application of OAMP, we assessed its hemolytic activity and cytotoxicity on blood cells across various concentrations (0, 1, 2, 4, 8, 16, 32 μM), building upon its MIC. Hemolytic activity of OAMP was evaluated using sheep red blood cells (RBCs) and chicken red blood cells (RBCs), and results are depicted in Figures 5A, B. The peptide OAMP exhibited increasing hemolytic rates with rising concentrations. When the concentration was less than 4μM, the hemolytic rate of sheep red blood cells was less than 20%, but that of chicken red blood cells exceeded 20% at 4μM. At the MIC concentration of OAMP (8μM), the hemolysis rates of both types of red blood cells exceeded 20%. When the incubation concentration of OAMP reached the highest detectable concentration of 32 μM, the hemolytic rates of both cell types exceeded 50%, with the hemolytic rate for sheep red blood cells reaching as high as 68.80%. Then, the cytotoxicity of OAMP is illustrated in Figure 5C. As the concentration of OAMP increases, the viability of RAW264.7 cells shows a decreasing trend. At a concentration of 32 μM, the cell viability reaches its lowest point at 53.84%. Within the range of OAMP MICs against CR-AB (MIC ≤ 8 µM), the viability of RAW264.7 cells remains above 60%.