Evaluation of the activity of antimicrobial peptides against bacterial vaginosis.

New drugs for the treatment of bacterial vaginosis (BV) are yet to be developed due to concerns that they may contribute to the increase in antibiotic resistance in BV. Antimicrobial peptides (AMPs) are one of the most promising options for next-generation antibiotics. In this study, we investigated the bacteriostatic activity of the AMPs Pexiganan, plectasin, melittin, and cathelicidin-DM against Gram-negative and Gram-positive bacteria both in vitro and in a mouse model of BV infection. The results showed that Pexiganan, melittin, and cathelicidin-DM had significant antibacterial activity against both Gram-negative and Gram-positive bacteria. AMPs have great potential for clinical application in the treatment of vaginitis, and this study provides an experimental basis for their use in the active immunoprophylaxis of BV.

From Data to Decisions: Leveraging Artificial Intelligence and Machine Learning in Combating Antimicrobial Resistance - a Comprehensive Review.

The emergence of drug-resistant bacteria poses a significant challenge to modern medicine. In response, Artificial Intelligence (AI) and Machine Learning (ML) algorithms have emerged as powerful tools for combating antimicrobial resistance (AMR). This review aims to explore the role of AI/ML in AMR management, with a focus on identifying pathogens, understanding resistance patterns, predicting treatment outcomes, and discovering new antibiotic agents. Recent advancements in AI/ML have enabled the efficient analysis of large datasets, facilitating the reliable prediction of AMR trends and treatment responses with minimal human intervention. ML algorithms can analyze genomic data to identify genetic markers associated with antibiotic resistance, enabling the development of targeted treatment strategies. Additionally, AI/ML techniques show promise in optimizing drug administration and developing alternatives to traditional antibiotics. By analyzing patient data and clinical outcomes, these technologies can assist healthcare providers in diagnosing infections, evaluating their severity, and selecting appropriate antimicrobial therapies. While integration of AI/ML in clinical settings is still in its infancy, advancements in data quality and algorithm development suggest that widespread clinical adoption is forthcoming. In conclusion, AI/ML holds significant promise for improving AMR management and treatment outcome.

Protein Network Alterations in G-CSF Treated Severe Congenital Neutropenia Patients and Beneficial Effects of Oral Health Intervention.

PURPOSE: Severe congenital neutropenia (SCN) is a raredisorder characterized by diminished neutrophil levels. Despite granulocytecolony-stimulating factor (G-CSF) treatment, SCN patients remain still prone tosevere infections, including periodontal disease-a significant oral healthrisk. This study investigates the host proteome and metaproteome in saliva andgingival crevicular fluid (GCF) of G-CSF-treated patients. EXPERIMENTAL DESIGN: We used label-free quantitative proteomics on saliva and GCF samples from SCN patients before (n = 10, mean age: 10.7 ± 6.6 years) and after a 6-month oral hygiene intervention (n = 9,mean age: 11.6 ± 5.27 years), and from 12 healthy controls. RESULTS: We quantified 894 proteins in saliva (648 human,246 bacterial) and 756 proteins in GCF (493 human, 263 bacterial). Predominant bacterial genera included Streptococcus, Veillonella, Selenomonas, Corynebacterium, Porphyromonas, and Prevotella. SCN patients showed reduced antimicrobial peptides (AMPs) and elevated complement proteins compared tohealthy controls. Oral hygiene intervention improved oral epithelial conditionsand reduced both AMPs and complement proteins. CONCLUSIONS AND CLINICAL RELEVANCE: SCN patients have aunique proteomic profile with reduced AMPs and increased complement proteins, contributing to infection susceptibility. Oral hygiene intervention not onlyimproved oral health in SCN patients but also offers potential overall therapeuticbenefits.

The topical application of Sphistin12-38 in combination with sponge spicules for the acne treatment.

We demonstrated for the first time that a marine-derived antimicrobial peptide (AMP), Sph12-38, exhibit high antimicrobial activity against P. acnes with a minimum bactericidal concentration (MBC) value of 7 µM. Meanwhile, Sph12-38 has no significant cytotoxicity to human keratinocytes (HKs) at its high concentration (33.5 µM). The topical application of sponge Haliclona sp. spicules (SHS) dramatically enhanced the skin penetration of Sph12-38 up to 40.9 ± 5.9% (p < 0.01), which was 6.1 ± 0.9-fold higher than that of Sph12-38 alone. Further, SHS resulted in the accumulation of most Sph12-38 in viable epidermis and dermis. Further, the combined use of Sph12-38 and SHS resulted in a cure rate of 100% for rabbit ear acne treatment in vivo for two weeks, while the one induced by other groups was 40%, 0% and 0% for SHS alone, Sph12-38 alone and control group, respectively. The strategy of combined using AMP and SHS can also be applied in a rational designed topical delivery system for the management of other deep infection of the skin. The effectiveness of SHS by itself on the treatment of acne was also demonstrated by clinical trials. After 14 days of treatment by 1% SHS gel. The number of skin lesions decreased by 51.4%.

Restoration of Type 17 immune signaling is not sufficient for protection during influenza-associated pulmonary aspergillosis.

Influenza-associated pulmonary aspergillosis (IAPA) is a severe complication of influenza infection that occurs in critically ill patients and results in higher mortality compared to influenza infection alone. Interleukin-17 (IL-17) and the Type 17 immune signaling pathway cytokine family are recognized for their pivotal role in fostering protective immunity against various pathogens. In this study, we investigate the role of IL-17 and Type 17 immune signaling components during IAPA. Wild-type mice were challenged with influenza A H1N1 (Flu) and then exposed to Aspergillus fumigatus ATCC42202 resting conidia on day 6 post-influenza infection, followed by the quantification of cytokines and chemokines at 48 hours post-fungal infection. Gene and protein expression levels revealed that IL-17 and Type 17 immune cytokines and antimicrobial peptides are downregulated during IAPA compared to mice singularly infected solely with A. fumigatus. Restoration of Type 17 immunity was not sufficient to provide protection against the increased fungal burden observed during IAPA. These findings contrast those observed during post-influenza bacterial super-infection, in which restoration of Type 17 immune signaling protects against exacerbation seen during super-infection. Our study highlights the need for future studies to understand the immune mechanisms that increase susceptibility to fungal infection.

Interaction of Acinetobacter sp. RIT 592 induces the production of broad-spectrum antibiotics in Exiguobacterium sp. RIT 594.

Antimicrobial resistance (AMR) is one of the most alarming global public health challenges of the 21st century. Over 3 million antimicrobial-resistant infections occur in the United States annually, with nearly 50,000 cases being fatal. Innovations in drug discovery methods and platforms are crucial to identify novel antibiotics to combat AMR. We present the isolation and characterization of potentially novel antibiotic lead compounds produced by the cross-feeding of two rhizosphere bacteria, Acinetobacter sp. RIT 592 and Exiguobacterium sp. RIT 594. We used solid-phase extraction (SPE) followed by liquid chromatography (LC) to enrich antibiotic extracts and subsequently mass spectrometry (MS) analysis of collected fractions for compound structure identification and characterization. The MS data were processed through the Global Natural Product Social Molecular Networking (GNPS) database. The supernatant from RIT 592 induced RIT 594 to produce a cocktail of antimicrobial compounds active against Gram-positive and negative bacteria. The GNPS analysis indicated compounds with known antimicrobial activity in the bioactive samples, including oligopeptides and their derivatives. This work emphasizes the utility of microbial community-based platforms to discover novel clinically relevant secondary metabolites. Future work includes further structural characterization and antibiotic activity evaluation of the individual compounds against pathogenic multidrug-resistant (MDR) bacteria.

Trematocine-derived antimicrobial peptides from the Antarctic fish Trematomus bernacchaii: potent antibacterial agents against ESKAPE pathogens.

Introduction: This study investigated the interaction with membrane mimetic systems (LUVs), bacterial membranes, the CD spectra, and the bactericidal activity of two designed trematocine mutants, named Trem-HK and Trem-HSK. Mutants were constructed from the scaffold of Trematocine (Trem), a natural 22-amino acid AMP from the Antarctic fish Trematomus bernacchii, aiming to increase their positive charge. Methods: The selectivity of the designed AMPs towards bacterial membranes was improved compared to Trematocine, verified by their interaction with different LUVs and their membranolytic activity. Additionally, their α-helical conformation was not influenced by the amino acid substitutions. Our findings revealed a significant enhancement in antibacterial efficacy against ESKAPE (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacteriaceae family) pathogens for both Trem-HK and Trem-HSK. Results: Firstly, we showed that the selectivity of the two new designed AMPs towards bacterial membranes was greatly improved compared to Trematocine, verifying their interaction with different LUVs and their membranolytic activity. We determined that their α-helical conformation was not influenced by the amino acid substitutions. We characterized the tested bacterial collection for resistance traits to different classes of antibiotics. The minimum inhibitory and bactericidal concentration (MIC and MBC) values of the ESKAPE collection were reduced by up to 80% compared to Trematocine. The bactericidal concentrations of Trematocine mutants showed important membranolytic action, evident by scanning electron microscopy, on all tested species. We further evaluated the cytotoxicity and hemolytic activity of the mutants. At 2.5 µM concentration, both mutants demonstrated low cytotoxicity and hemolysis, indicating selectivity towards bacterial cells. However, these effects increased at higher concentrations. Discussion: Assessment of in vivo toxicity using the Galleria mellonella model revealed no adverse effects in larvae treated with both mutants, even at concentrations up to 20 times higher than the lowest MIC observed for Acinetobacter baumannii, suggesting a high potential safety profile for the mutants. This study highlights the significant improvement in antibacterial efficacy achieved by increasing the positive charge of Trem-HK and Trem-HSK. This improvement was reached at the cost of reduced biocompatibility. Further research is necessary to optimize the balance between efficacy and safety for these promising AMPs.

Promising anti-inflammatory activity of a novel designed anti-microbial peptide for wound healing.

Chronic wounds can develop as a result of prolonged inflammation during the healing process, which can happen due to bacterial infection. Therefore, preventing infection and controlling inflammation can accelerate wound healing. Antimicrobial peptides have different protective properties in addition to antimicrobial activity. Some of these activities include the stimulation of cytokine or chemokine synthesis, the facilitation of chemotaxis and cell proliferation, the acceleration of cell proliferation, the induction of anti-inflammatory responses, and the promotion of wound repair. This study aimed to assess the wound healing potential of a novel in silico-designed antimicrobial peptide. Then, its anti-inflammatory activity was investigated by measuring the level of tumor necrosis factor-α (TNF-α) and transforming growth factor beta (TGF-ß) as indicators of the wound healing process. In addition, the influence of the peptide on cell migration was evaluated by a scratch test on human dermal fibroblasts (HDF) and HaCaT cells as a human epidermal keratinocyte cell line. The results showed that our new peptide could act well in inhibiting TNF-α over-secretion while increasing the expression of TGF-ß as an anti-inflammatory factor. This peptide showed a significant potential to stimulate HDF and HaCaT cell migration and proliferation. Therefore, using this peptide as an anti-inflammatory component of wound dressings may be promising.

Assessing the antiviral activity of antimicrobial peptides Caerin1.1 against PRRSV in Vitro and in Vivo.

The Porcine reproductive and respiratory syndrome (PRRS) causes severe financial losses to the global swine industry. Due to continuous virus evolution, the protection against the PRRS provided by current vaccines is limited. In order to find new antiviral strategies, this study investigated the antiviral potential of antimicrobial peptides (AMPs) against PRRSV. Given the diversity of PRRSV strains and the limited effectiveness of existing vaccines in controlling PRRSV, this study evaluated the inhibitory effects of KLAK, Cecropin B, Piscidin1, and Caerin1.1 on 3 strains of PRRSV (lineage 5 classical strain, lineage 8 highly pathogenic strain, and lineage 1 NADC30-like strain). Caerin1.1 exhibited significant dose-dependent antiviral activity, with an effective concentration (EC50) of 7.5 µM. Caerin1.1 effectively inhibited PRRSV replication when added before or in early infection but showed reduced effectiveness when added in late infection, indicating its potential involvement in targeting early transcription mechanisms of viral RNA polymerase and significantly upregulating cytokine gene expression. In the NADC30 strain-based animal infection model, Caerin1.1 treatment significantly reduced lung viral loads and inflammation in the lungs of PRRSV-infected pigs, with a mortality rate of 0 % (0/5) in the treated group compared to 66.67 % (4/6) in the untreated group, indicating a reduction in the mortality rate. Additionally, compared with the untreated group, the Caerin1.1-treated group showed significant improvements, such as lighter fever, more daily weight gain, less clinical symptoms, less viral load in blood, and less virus oral shedding (P < 0.05). These findings reveal the potential of antimicrobial peptides as PRRSV therapeutic agents and suggest that Caerin1.1 is a promising candidate for a novel anti-PRRSV drug.

Mining human microbiomes reveals an untapped source of peptide antibiotics.

Drug-resistant bacteria are outpacing traditional antibiotic discovery efforts. Here, we computationally screened 444,054 previously reported putative small protein families from 1,773 human metagenomes for antimicrobial properties, identifying 323 candidates encoded in small open reading frames (smORFs). To test our computational predictions, 78 peptides were synthesized and screened for antimicrobial activity in vitro, with 70.5% displaying antimicrobial activity. As these compounds were different compared with previously reported antimicrobial peptides, we termed them smORF-encoded peptides (SEPs). SEPs killed bacteria by targeting their membrane, synergizing with each other, and modulating gut commensals, indicating a potential role in reconfiguring microbiome communities in addition to counteracting pathogens. The lead candidates were anti-infective in both murine skin abscess and deep thigh infection models. Notably, prevotellin-2 from Prevotella copri presented activity comparable to the commonly used antibiotic polymyxin B. Our report supports the existence of hundreds of antimicrobials in the human microbiome amenable to clinical translation.

Engineered extracellular vesicles coated with an antimicrobial peptide for advanced control of bacterial sepsis.

Alarming sepsis-related mortality rates present significant challenges to healthcare services globally. Despite advances made in the field, there is still an urgent need to develop innovative approaches that could improve survival rates and reduce the overall cost of treatment for sepsis patients. Therefore, this study aimed to develop a novel multifunctional therapeutic agent for advanced control of bacterial sepsis. Extracellular vesicles (EVs) isolated from lipopolysaccharide (LPS) induced HepG2 (hepatocellular carcinoma cells) (iEV) displayed an average particle size of 171.63 ± 2.77 nm, a poly dispersion index (PDI) of 0.32 ± 0.0, and a zeta potential (ZP) of -11.87 ± 0.18 mV. Compared to HepG2 EV, LPS induction significantly increases the EV protein concentration, PDI and ZP, reduces the average size and promotes cell proliferation and cytoprotective effects of the isolated EVs (iEVs) against LPS-induced cytotoxicity. Coating of iEV with a cationic antimicrobial peptide (AMP) to form PC-iEV slightly changed their physical properties and shifted their surface charge toward neutral values. This modification improved the antibacterial activity (2-fold lower minimum bactericidal concentration [MBC] values) and biocompatibility of the conjugated peptide while maintaining iEV cytoprotective and anti-inflammatory activities. Our findings indicate the superior anti-inflammatory and antibacterial dual activity of PC-iEV against pathogens associated with sepsis.

Synergistic action of antimicrobial peptides and antibiotics: current understanding and future directions.

Antibiotic resistance is a growing global problem that requires innovative therapeutic approaches and strategies for administering antibiotics. One promising approach is combination therapy, in which two or more drugs are combined to combat an infection. Along this line, the combination of antimicrobial peptides (AMPs) with conventional antibiotics has gained attention mainly due to the complementary mechanisms of action of AMPs and conventional antibiotics. In this article, we review both in vitro and in vivo studies that explore the synergy between AMPs and antibiotics. We highlight several mechanisms through which synergy is observed in in vitro experiments, including increasing membrane permeability, disrupting biofilms, directly potentiating antibiotic efficacy, and inhibiting resistance development. Moreover, in vivo studies reveal additional mechanisms such as enhanced/modulated immune responses, reduced inflammation, and improved tissue regeneration. Together, the current literature demonstrates that AMP-antibiotic combinations can substantially enhance efficacy of antibiotic therapies, including therapies against resistant bacteria, which represents a valuable enhancement to current antimicrobial strategies.

In Silico and In vitro Evaluations of the Antibacterial Activities of HIV-1 Nef Peptides against Pseudomonas aeruginosa.

One of the burning issues facing healthcare organizations is multidrug-resistant (MDR) bacteria. P. aeruginosa is an MDR opportunistic bacterium responsible for nosocomial and fatal infections in immunosuppressed individuals. According to previous studies, efflux pump activity and biofilm formation are the most common resistance mechanisms in P. aeruginosa. The aim of this study was to propose new antimicrobial peptides (AMPs) that target P. aeruginosa and can effectively address these resistance mechanisms through in silico and in vitro assessments. Since AMPs are an attractive alternative to antibiotics, in vitro experiments were carried out along with bioinformatics analyses on 19 Nef peptides (derived from the HIV-1 Nef protein) in the current study. Several servers, including Dbaasps, Antibp2, CLASSAMP2, ToxinPred, dPABBs and ProtParam were used to predict Nef peptides as AMPs. To evaluate the binding affinities, a molecular docking analysis was performed with the HADDOCK web server for all Nef peptide models against two effective proteins of P. aeruginosa (MexB and PqsR) that play a role in efflux and quorum sensing. Moreover, the antibacterial and antibiofilm activity of the Nef peptides was investigated in a resistant strain of P. aeruginosa. The results of molecular docking revealed that all Nef peptides have a significant binding affinity to the abovementioned proteins. Nef-Peptide-19 has the highest affinity to the active sites of MexB and PqsR with the HADDOCK scores of -136.1 ± 1.7 and -129.4 ± 2, respectively. According to the results of in vitro evaluation, Nef peptide 19 showed remarked activity against P. aeruginosa with minimum inhibitory and bactericidal concen-trations (MIC and MBC) of 10 µM and 20 µM, respectively. In addition, biofilm inhibitory activity was observed at a concentration of 20 µM. Finally, Nef peptide 19 is proposed as a new AMP against P. aeruginosa.

Venom: A Promising Avenue for Antimicrobial Therapeutics.

Venom in medicine is well documented in the chronicles of ancient Greece and the Roman Empire and persisted into the Renaissance and even into the modern era. Venoms were not always associated with detrimental consequences. Since ancient times, the curative capacity of venom has been recognized, portraying venom as a metaphor for pharmacy and medicine. Venom proteins and peptides' antimicrobial potential has not undergone systematic exploration despite the huge literature on natural antimicrobials. In light of the escalating challenge of antimicrobial resistance and the diminishing effectiveness of antibiotics, there is a pressing need for innovative antimicrobials capable of effectively addressing illnesses caused by multidrug-resistant microorganisms. This review adds to our understanding of the effectiveness of different venom components against a host of pathogenic microorganisms. The aim is to illuminate the various antimicrobials present in venom and venom peptides, thereby emphasizing the unexplored medicinal potential for antimicrobial properties. We have presented a concise summary of the molecular examination of the venom peptides' functioning processes, as well as the current clinical and preclinical progress of venom antimicrobial peptides.

High Therapeutic Index α-Helical AMPs and Their Therapeutic Potential on Bacterial Lung and Skin Wound Infections.

Antimicrobial peptides (AMPs) possess strong antibacterial activity and low drug resistance, making them ideal candidates for bactericidal drugs for addressing the issue of traditional antibiotic resistance. In this study, a template (G(XXKK)nI, G = Gly; X = Leu, Ile, Phe, or Trp; n = 2, 3, or 4; K = Lys; I = Ile.) was employed for the devised of a variety of novel α-helical AMPs with a high therapeutic index. The AMP with the highest therapeutic index, WK2, was ultimately chosen following a thorough screening process. It demonstrates broad-spectrum and potent activity against both standard and multidrug-resistant bacteria, while also showing low hemolysis and rapid and efficient time-kill kinetics. Additionally, WK2 exhibits excellent efficacy in treating mouse models of Klebsiella pneumonia-induced lung infections and methicillin-resistant Staphylococcus aureus (MRSA)-induced skin wound infections while demonstrating good safety profiles in vivo. In conclusion, the template-based design methodology for novel AMPs with high therapeutic indices offers new insights into addressing antibiotic resistance problems. WK2 represents a promising antimicrobial agent.

Cocktail Approach with Polyserotonin Nanoparticles and Peptides for Treatment of Streptococcus mutans.

Dental plaque, formed by a Streptococcus mutans biofilm, is a major contributor to cavity formation. While antimicrobial strategies exist, the growing risk of antibiotic resistance necessitates alternative therapeutic solutions. Polyserotonin nanoparticles (PSeNPs), recently recognized for their photothermal property and promising biomedical applications, open up a new avenue for antimicrobial use. Here, we introduced a UV-initiated synthetic route for PSeNPs with improved yield. Using these PSeNPs, a cocktail treatment to reduce the viability of this cavity-causing bacteria was developed. This cocktail comprises an S. mutans-targeting antimicrobial peptide (GH12), an intraspecies competence-stimulating peptide that triggers altruistic cell death in S. mutans, and laser-activated heating of PSeNPs. The "peptide + PSeNP + laser" combination effectively inhibits S. mutans growth in both planktonic and biofilm states. Moreover, the cocktail approach remains effective in reducing the viability of S. mutans in a more virulent dual-species biofilm with Candida albicans. Overall, our results reinforce the utility of a multipronged therapeutic strategy to reduce cariogenic bacteria in the complex model oral biofilm.

Genomic Sequence of Streptococcus salivarius MDI13 and Latilactobacillus sakei MEI5: Two Promising Probiotic Strains Isolated from European Hakes (Merluccius merluccius, L.).

Frequently, diseases in aquaculture have been fought indiscriminately with the use of antibiotics, which has led to the development and dissemination of (multiple) antibiotic resistances in bacteria. Consequently, it is necessary to look for alternative and complementary approaches to chemotheraphy that are safe for humans, animals, and the environment, such as the use of probiotics in fish farming. The objective of this work was the Whole-Genome Sequencing (WGS) and bioinformatic and functional analyses of S. salivarius MDI13 and L. sakei MEI5, two LAB strains isolated from the gut of commercial European hakes (M. merluccius, L.) caught in the Northeast Atlantic Ocean. The WGS and bioinformatic and functional analyses confirmed the lack of transferable antibiotic resistance genes, the lack of virulence and pathogenicity issues, and their potentially probiotic characteristics. Specifically, genes involved in adhesion and aggregation, vitamin biosynthesis, and amino acid metabolism were detected in both strains. In addition, genes related to lactic acid production, active metabolism, and/or adaptation to stress and adverse conditions in the host gastrointestinal tract were detected in L. sakei MEI5. Moreover, a gene cluster encoding three bacteriocins (SlvV, BlpK, and BlpE) was identified in the genome of S. salivarius MDI13. The in vitro-synthesized bacteriocin BlpK showed antimicrobial activity against the ichthyopathogens Lc. garvieae and S. parauberis. Altogether, our results suggest that S. salivarius MDI13 and L. sakei MEI5 have a strong potential as probiotics to prevent fish diseases in aquaculture as an appropriate alternative/complementary strategy to the use of antibiotics.

Structure-aware deep learning model for peptide toxicity prediction.

Antimicrobial resistance is a critical public health concern, necessitating the exploration of alternative treatments. While antimicrobial peptides (AMPs) show promise, assessing their toxicity using traditional wet lab methods is both time-consuming and costly. We introduce tAMPer, a novel multi-modal deep learning model designed to predict peptide toxicity by integrating the underlying amino acid sequence composition and the three-dimensional structure of peptides. tAMPer adopts a graph-based representation for peptides, encoding ColabFold-predicted structures, where nodes represent amino acids and edges represent spatial interactions. Structural features are extracted using graph neural networks, and recurrent neural networks capture sequential dependencies. tAMPer's performance was assessed on a publicly available protein toxicity benchmark and an AMP hemolysis data we generated. On the latter, tAMPer achieves an F1-score of 68.7%, outperforming the second-best method by 23.4%. On the protein benchmark, tAMPer exhibited an improvement of over 3.0% in the F1-score compared to current state-of-the-art methods. We anticipate tAMPer to accelerate AMP discovery and development by reducing the reliance on laborious toxicity screening experiments.

Discovery of AMPs from random peptides via deep learning-based model and biological activity validation.

The ample peptide field is the best source for discovering clinically available novel antimicrobial peptides (AMPs) to address emerging drug resistance. However, discovering novel AMPs is complex and expensive, representing a major challenge. Recent advances in artificial intelligence (AI) have significantly improved the efficiency of identifying antimicrobial peptides from large libraries, whereas using random peptides as negative data increases the difficulty of discovering antimicrobial peptides from random peptides using discriminative models. In this study, we constructed three multi-discriminator models using deep learning and successfully screened twelve AMPs from a library of 30,000 random peptides. three candidate peptides (P2, P11, and P12) were screened by antimicrobial experiments, and further experiments showed that they not only possessed excellent antimicrobial activity but also had extremely low hemolytic activity. Mechanistic studies showed that these peptides exerted their bactericidal effects through membrane disruption, thus reducing the possibility of bacterial resistance. Notably, peptide 12 (P12) showed significant efficacy in a mouse model of Staphylococcus aureus wound infection with low toxicity to major organs at the highest tested dose (400 mg/kg). These results suggest deep learning-based multi-discriminator models can identify AMPs from random peptides with potential clinical applications.