Pseudomonas asiaticais an emerging opportunistic pathogen with a broad host range. Current evidence suggests that some isolates exhibit multidrug resistance, which may complicate treatment. In this study, a multidrug-resistant P. asiatica strain MD9 was isolated from aquaculture water. We aimed to characterize its complete genome sequence and investigate the role of its conjugative megaplasmid pMD9A in the horizontal transfer of antibiotic resistance genes. The genome of MD9 consists of one circular chromosome (5,956,782 bp, with a G + C content of 62.5%) and one circular megaplasmid, pMD9A (455,169 bp, with a G + C content of 56.5%). Genome annotation identified 65 antibiotic resistance genes and 148 putative virulence factor-encoding genes in the MD9 genome. The megaplasmid pMD9A carries 29 antibiotic resistance genes conferring resistance to β-lactams, chloramphenicol/florfenicol, aminoglycosides, and macrolides. A class 1 integron (intI1) and multiple autonomous conjugative transfer elements were identified in pMD9A. Conjugation experiments demonstrated that the β-lactam resistance gene blaOXA-246 could be horizontally transferred from the donor MD9 strain to the recipient Escherichia coli 25DN strain. The megaplasmid pMD9A not only carries a broad array of antibiotic resistance genes, but also facilitates their horizontal spread among environmental bacteria, thereby potentially contributing to the dissemination of multidrug-resistant bacteria.

The objective of this study was to analyze longitudinal trends in the detection rates of Carbapenem-resistant Gram-negative bacilli (CRGNB) in China from 2014 to 2024, and to forecast their evolution over the subsequent five years. Surveillance data were extracted from the annual reports of the China Antimicrobial Resistance Surveillance System (CARSS). In the period between 2014 and 2024, carbapenem-resistant Acinetobacter baumannii (CRAB) was identified with the highest frequency (range: 52.20%-60.00%), followed by carbapenem-resistant Pseudomonas aeruginosa (CRPA) (16.00%-25.60%), carbapenem-resistant Klebsiella pneumoniae (CRKP) (6.40%-11.30%), and carbapenem-resistant Escherichia coli (CREC) (1.50%-1.90%). Joinpoint analysis revealed a significant overall decreasing trend for CRPA (Average Annual Percentage Change, AAPC = -4.64%) and CRAB (AAPC = -1.05%). CRKP demonstrated a significant overall increase (AAPC = 4.29%), characterized by a sharp rise from 2014 to 2019 (Annual Percentage Change, APC = 10.73%, p < 0.01) followed by a plateau and nonsignificant decline (APC = -1.77%, p = 0.31). CREC exhibited a non-significant decreasing trend (AAPC = -1.42%, p = 0.16). The Autoregressive Integrated Moving Average (ARIMA) models, which were fitted to the data, demonstrated excellent predictive accuracy. Projections for 2025 indicate detection rates of 1.61% for CREC, 9.63% for CRKP, 14.25% for CRPA, and 52.39% for CRAB. Projections indicate a persistent downward trajectory over the ensuing five-year period, thereby substantiating the ongoing necessity for antimicrobial stewardship and surveillance initiatives.

Pseudomonas aeruginosa is a notorious opportunistic pathogen constituting a major cause of health care-associated infections worldwide. Intrinsic resistance to multiple antibiotics, with acquired resistance mechanisms and hypervirulence make it a critical target for genomic investigations. In this study, we investigated AMR determinants and virulence factors using WGS among five extensively drug-resistant (XDR) clinical P. aeruginosa isolates. XDR P. aeruginosa isolates isolated from bile, broncho-alveolar lavage and wound swabs were included in the study. Initially, the isolates were tested for the carbapenemase phenotype and genotype using modified Hodge test and PCR, respectively, followed by WGS. All isolates exhibited resistance to >50% of the antibiotics tested. Conventional PCR analysis identified blaNDM-1, blaOXA-48, and blaVIM invariably in the test isolates. Tested isolates were assigned sequence types as ST357 (n = 2), ST773, ST1047, and ST3043. WGS identified multiple resistant genes including blaOXA-488 and blaVIM-11(P2356), blaOXA-10, blaVIM-2, blaOXA-395 and blaOXA50 (PM381), blaVEB-9, blaOXA-50 and blaNDM-1 (PR90), blaOXA-485 (PR126), and blaOXA-50 (PR64). Notably, all the isolates were carrying different variants of blaPDC. The genomes were also associated with the virulence determinants, with genes encoding alginate and biofilm synthesis, quorum sensing, secretion systems, secondary metabolites, lipopolysaccharides, exotoxins, pili, and flagellar structure. All clinical P. aeruginosa isolates exhibited extensive resistance and virulence gene repertoires, also enriched with efflux pumps. Sequence types ST3043 and ST1047 are reported for the first time in India, underscoring the need for continuous genomic surveillance to monitor the emergence and dissemination of high-risk clones.

Healthcare-associated infections (HAIs) and antimicrobial resistance (AMR) are significant problems, especially in intensive care units (ICUs). We evaluated HAI rates, causative pathogens, resistance profiles, and antibiotic consumption retrospectively in a tertiary ICU across three 2-year periods-pre-pandemic period (PreP), pandemic period (PanP), and post-pandemic period (PostP)-to assess the impact of the COVID-19 pandemic. Antibiotic consumption was expressed as defined daily dose (DDD) per 1,000 patient days. The highest infection incidence rate and density were in PanP. Despite lower invasive device use, PanP had a higher rate of catheter-related bloodstream infections (BSIs). The most frequent pathogen across periods was Acinetobacter spp., although its rate declined in PostP. Gram-negative bacteria resistance rates to piperacillin-tazobactam, carbapenem, and aminoglycoside were highest in PanP. Resistant pathogen and resistant infection rates were lower in PostP, and resistant infections were less frequent in PostP than in PanP (OR: 0.371; 95% CI: 0.216-0.637; p < 0.001). Healthcare-associated pneumonia (HAP) was associated with a higher risk of resistance compared with BSI (OR: 3.467; 95% CI: 2.139-5.621; p < 0.001). Total antibiotic consumption remained comparable between PreP and PanP but increased in PostP. During the PanP, carbapenem use increased while piperacillin-tazobactam consumption decreased, whereas glycopeptide use increased and reached its highest level in PostP. Pandemic-related disruptions in the health care system may have contributed to increases in HAI and AMR rates, whereas improvements in the PostP may have supported recovery. Continuous surveillance, infection prevention measures, and antimicrobial stewardship programs are essential to mitigate the impact of future health care crises.

ICESan95 is a serine integrase integrative and conjugative element (ICE) that targets methyltransferase genes for integration. The mosaic ICE ICESag084 contains ICESan95 and partial fragments from two distinct genes (snf2 and hsdM), suggesting a recombination event. This study aimed to experimentally investigate the recombination of multiple ICESan95, leading to the formation of a mosaic ICE resembling ICESag084. Using conjugation assays, we tracked this evolutionary process, confirmed the structures via whole-genome sequencing, and assessed genomic stability through serial passages. Analysis showed that ICESag084 comprises an ICESan95-like ICE, a tyrosine integrase element, and a partial ICESpy009 sequence. The ICESan95-like ICE was flanked by truncated snf2 and hsdM genes, indicating ancestral cointegration of two ICESan95-like ICEs followed by excision and sequence capture. Experiments confirmed that tandem integration of ICESan95 into adjacent snf2 (from an ICESpy009-like element) and hsdM(from the tyrosine integrase element) sites facilitated recombinational excision. This event deleted an ∼36.6-kb intervening fragment, yielding an ∼86.5-kb circularized intermediate and generating an ∼81.0-kb mosaic ICE highly similar to ICESag084. The mosaic ICE was conjugatively transferable and remained stable for ≥60 generations, while the circularized ICE was lost within 20 generations. Thus, tandem integration and recombinational excision drive the emergence and spread of mosaic ICEs. This mechanism facilitates the modular assembly of multiple mobile genetic elements, including ICESpy009, ICESan95, and a tyrosine integrase element. This process creates novel resistance islands with expanded traits, providing a key pathway for long-term adaptive evolution in bacterial pathogens.

The emergence of Escherichia coli sequence type (ST131), especially its H30Rx subclade, has reshaped the epidemiology of urinary tract infections (UTIs) by combining multidrug resistance (MDR), extended-spectrum β-lactamase (ESBL) production, and biofilm formation. Clinical E. coli isolates were genotyped for ST131 and H30Rx using PCR. Antimicrobial susceptibility, ESBL activity, biofilm formation, and plasmid carriage were assessed through standard protocols. Statistical tests were applied to explore associations between clonal types and resistance traits. Nearly half of the isolates (46.3%) were identified as ST131, with 78.9% belonging to H30Rx. Overall, eight isolates (19.5%) were ESBL producers. ESBL production was significantly higher in ST131 (36.8%) compared with non-ST131 (4.5%; p = 0.004). Plasmid carriage was universal in ST131 (100%) and more frequent than in non-ST131 (86.4%; p = 0.029). MDR was observed in almost 90% of ST131 isolates. Biofilm formation was widespread (97.5%), with all ST131 strains positive. Resistance was greatest against fluoroquinolones and third-generation cephalosporins, while fosfomycin, nitrofurantoin, and amikacin retained high activity with low resistance rates. The predominance of ST131, particularly H30Rx, highlights its role as a high-risk clone driving ESBL production, plasmid-mediated resistance, and biofilm-associated persistence in UTIs. Given the rising resistance to commonly used antibiotics, fosfomycin, nitrofurantoin, and amikacin remain promising therapeutic options. Key limitations include a modest sample size (n = 41 E. coli isolates), the absence of genotypic ESBL characterization, and the use of disk diffusion for fosfomycin and nitrofurantoin without MIC confirmation. Continuous surveillance, molecular characterization, and strengthened antimicrobial stewardship are essential to control the spread of this lineage and preserve treatment efficacy.

Human infections with non-Helicobacter pylori Helicobacter species (NHPH) are rare. The diagnosis of such infection relies strongly on histological methods, as these bacteria cannot be easily cultured. We describe microbiological analyses that were performed on three cases of infection reported between 2020 and 2023, for which molecular biology was used to identify the species and facilitate the investigation of markers of resistance to macrolides and fluoroquinolones. 16S rRNA sequencing identified two Helicobacter suis strains and one Helicobacter heilmannii sensu strictostrain. Sequencing of 23S rRNA and quinolone resistance determining region of the gyrA gene, classically described as carrying mutations associated with resistance to macrolide and levofloxacin, respectively, did not reveal any mutations. RIDA[®]GENE H. pylori, the Allplex™ H. pylori and ClariR Assay and Amplidiag[®] H. pylori+ClariR cross-react with these NHPH strains. Any case of NHPH infection diagnosed by histology should be confirmed by molecular biology in specialized laboratories to identify the species and investigate the resistome.

Since their first identification in Türkiye in 2001, OXA-48-like carbapenemases have posed diagnostic challenges due to variant-specific phenotypic and resistance profiles. We investigated the distribution of OXA-48 variants and their association with antimicrobial susceptibility, carbapenem minimum inhibitory concentrations (MICs), phenotypic detection performance, and single- or dual-carbapenemase production. A total of 703 clinical carbapenem-resistant Enterobacterales isolates recovered over 5 years were included. Identification and antimicrobial susceptibility testing were performed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) and an automated system; carbapenem MICs by broth microdilution; carbapenemase genes by multiplex qPCR; and OXA-48 variants by sequence analysis. CNPt-direct and mCIM were performed for OXA-48 variants. OXA-48-like, NDM, and KPC carbapenemases were detected in 52.3%, 11.7%, and 11.5% of isolates, respectively, with co-production observed in 11.7% of OXA-48-like-positive isolates. The most common OXA-48 variants were OXA-48/245 (39.1%), OXA-232 (36.4%), and OXA-181 (17.6%), while OXA-244, OXA-162, and OXA-1200 were less frequent. This study represents the first report of the OXA-1200 variant from Türkiye. OXA-48/245 was the variant most commonly co-produced with other carbapenemases, whereas OXA-244 predominated among CNPt-direct-negative isolates. Susceptibility to meropenem and imipenem among OXA-48-like producers was 31% and 39.1%, respectively, with higher carbapenem MIC50 values observed for OXA-181, OXA-232, and OXA-48/245 compared to OXA-244. In K. pneumoniae, meropenem resistance rates were higher with OXA-48/245, OXA-232, and OXA-181 than with other variants. MIC50 values of carbapenem were higher in dual carbapenemase producers compared to single carbapenemase producers. Our findings show that OXA-48 variants significantly impact resistance profiles, MIC values, and the sensitivity of phenotypic tests and detection of co-produced carbapenemases. Understanding their regional distribution is crucial for targeted prevention strategies.

This study evaluated a 6-month, on-farm antimicrobial stewardship model integrating veterinary oversight, appropriate antimicrobial use, enhanced biosecurity, and an all-in/all-out system on five Korean pig farms lacking prior oversight. Antimicrobial usage and resistance of Escherichia coli were compared before and after implementation. The number of antimicrobial prescriptions decreased significantly following the implementation of the model (from 15.2 ± 9.2 to 7.2 ± 4.7). After model implementation, prescriptions for suckling piglets decreased from 2.2 ± 1.6 to 0.4 ± 0.9, and for weaned piglets from 4.4 ± 2.4 to 1.8 ± 1.1. Overall antimicrobial resistance decreased following model implementation, with the largest decreases were observed in ciprofloxacin (from 57.3% to 21.9%), nalidixic acid (from 60.0% to 34.2%), and cefazolin (from 92.0% to 64.4%). The multidrug-resistance rates in eight antimicrobial classes decreased significantly across all stages (from 28.0% to 4.1%), with decreases for suckling piglets (from 33.3% to 7.1%), weaned piglets (from 40.0% to 0.0%), growers (from 26.7% to 0.0%), finishers (from 26.7% to 14.3%), and sows (from 13.3% to 0.0%). The model effectively reduced both antimicrobial use and resistance, supporting its value for antimicrobial stewardship in pig production.

Antimicrobial resistance (AMR) represents a critical global health crisis, driving increased mortality, treatment failure, and economic burden. Artificial intelligence (AI) offers transformative potential to counter this threat by enhancing detection, diagnostics, and therapeutic precision. This narrative review synthesizes recent advances in AI-based approaches for AMR prediction, antimicrobial discovery, and clinical decision support, drawing on representative peer-reviewed studies published between January 1, 2015, and April 24, 2026. Models such as Deeparg-LS, XGBoost, and vision transformers achieved remarkable predictive accuracy using genomic, spectroscopic, and clinical data (AUC > 0.90; sensitivity/specificity >95%). AI-driven clinical decision support systems reduced antibiotic mismatches by up to 67%, while generative algorithms accelerated antimicrobial peptide discovery with 76% validation success. Deep learning frameworks improved metagenomic resistance profiling, and microscopy-based diagnostics shortened antimicrobial susceptibility testing by 50-70%. However, major challenges persist, including dataset heterogeneity, computational intensity, limited model transferability, and ethical concerns related to data privacy, bias, and interpretability. Emerging strategies such as explainable AI and federated learning show promise in addressing these issues. Overall, AI stands as a pivotal enabler in the fight against AMR, with future progress hinging on interdisciplinary collaboration, standardized validation, and responsible integration into clinical practice.