Gordy, Garrett G.

ABSTRACT The increasing prevalence of carbapenem-resistant Klebsiella pneumoniae represents a serious global health challenge. Rapid and accurate detection methods are essential to inform early and correct use of antimicrobial therapy to idealize patient outcomes and limit the spread of resistant isolates. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) has been evaluated for resistance detection, but its clinical application remains limited. In this study, MALDI-TOF MS was integrated with machine learning (ML) to improve detection of Klebsiella pneumoniae carbapenemase (KPC) among K. pneumoniae isolates harboring KPC carbapenemase. Ensemble learning strategies were applied to 435 clinical isolates from different geographic locations (Spain, United States, South America, and Asia), with validation performed using external data sets. By constructing two peak matrices and applying four ML algorithms, as well as combinations of two and three in ensemble models, 92 different classifiers were tested. Ensemble combinations increased the specificity of classifiers to over 95%, while sensitivity reached 72%, being significantly higher than that of the MALDI Biotyper KPC module. Low sensitivity may be affected by technical variability during spectral acquisition. The first annotated MALDI-TOF MS spectrum for K. pneumoniae by in silico prediction of protein masses was developed to enable peak identification, including peaks potentially related to antimicrobial resistance. Overall, the results of this study show that combining MALDI-TOF MS with ensemble learning can enhance KPC detection performance. IMPORTANCE Rapid and accurate detection of Klebsiella pneumoniae carbapenemase (KPC)-producing Klebsiella pneumoniae informs early and correct use of antimicrobial therapy to idealize patient outcomes and limit the spread of antimicrobial resistance. In this study, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) and ensemble machine learning were employed to detect KPC-producing isolates, using diverse isolate collections, obtaining excellent specificity. An annotated K. pneumoniae MALDI-TOF MS profile is provided with peak annotation, providing a resource for peak analysis.

Three previously uncharacterized bacterial species, designated strains BD586 T , BD613 T and BD626 T , isolated from human clinical specimens, were identified at Mayo Clinic, Rochester, Minnesota, USA. Initial identification efforts using matrix-assisted laser desorption ionization-time-of-flight MS and partial 16S rRNA gene sequencing proved inconclusive. Comprehensive analysis involving phenotypic characterization, biochemical assays and whole-genome sequencing was undertaken. The isolates were Gram-negative, motile, facultatively anaerobic rods, occurring singly, in pairs and in short chains; BD613 T additionally formed small aggregates. The isolates tested positive for catalase and negative for oxidase. Their colonies appeared smooth, white, opaque and non-haemolytic. Growth was observed at 35 °C under aerobic, anaerobic and CO₂-enriched conditions, as well as in media with NaCl concentrations up to 10% and at pH 7-9. Phylogenetic relationships were inferred from core gene alignments, average nucleotide identity and digital DNA–DNA hybridization comparisons. Results confirmed the placement of BD586 T , BD613 T and BD626 T within the recently established Dryocola genus, while also indicating their novelty as distinct species. The major cellular fatty acids were C 16:0 and C 17:0 cyclo. Based on these findings, a formal description of three new species, Dryocola mayonis sp. nov. (type strain BD586 T =TSD 474 T , =NCTC 15089 T , =DSM 119465 T ), Dryocola sharpae sp. nov. (type strain BD613 T =TSD 475 T , =NCTC 15090 T , =DSM 119466 T ) and Dryocola baronae sp. nov. (type strain BD626 T =TSD 476 T , =NCTC 15091 T , =DSM 119479 T ), is proposed, and the genus description of Dryocola is emended to refine its genomic, morphological and chemotaxonomic boundaries.

A novel Corynebacterium species, strain BD556T, isolated from blood, was identified at Mayo Clinic, Rochester, MN, USA. After failing definitive identification using MALDI-ToF MS and partial 16S rRNA gene sequencing, BD556T was characterized using a polyphasic approach, including phenotypic, biochemical and whole-genome sequencing methods. BD556T was a Gram-positive rod with clubbed ends, facultatively anaerobic, catalase-positive, oxidase-negative and non-motile. Colonies were white, opaque and non-haemolytic with halo-like edges. BD556T grew at 35 °C in room air, with CO2 and under anaerobic conditions. BD556T grew well in 0 and 6% NaCl and weakly in 10% NaCl. The genome size was 2 349 779 bp with a G+C content of 60.39%. Phylogenetic analysis using 16S rRNA gene sequence analysis, average nucleotide identity and digital DNA–DNA hybridization between the genome of BD556T and the closest type strains from the Type Strain Genome Server database yielded separation values well beyond those required for species delineation. Chemotaxonomic analyses of BD556T revealed ribose, arabinose and galactose as whole-cell sugars and an A1γ meso-diaminopimelic acid-direct peptidoglycan type. The major cellular fatty acids were C15 : 0 (21.0%), C16 : 0 (14.8%), C17 : 1 ω9c (26.2%), C17 : 0 (13.3%) and C18 : 1 ω9c (18.3%). Polar lipids included diphosphatidylglycerol, phosphatidylglycerol and unidentified glycolipids and phospholipids. BD556T also contained mycolic acids (32–36 carbons) typical of corynebacteria. The respiratory quinones were dominated by MK-8(H2) (71.2%) and MK-9(H2) (25.9%), with smaller amounts of MK-7(H2) and MK-10(H2). The results presented support the tenet that BD556T (=TSD 427T=NCTC 15078T) is a novel species for which the name Corynebacterium mayonis sp. nov. is proposed.

Rapid detection of Klebsiella pneumoniae carbapenemase (KPC) in the Klebsiella species is desirable. The MALDI Biotyper® MBT Subtyping Module (Bruker Daltonics) uses an algorithm that detects a peak at ~11,109 m/z corresponding to a protein encoded by the p019 gene to detect KPC simultaneously with organism identification by a matrix-assisted laser desorption ionization–time-of-flight mass spectrometry (MALDI-ToF MS). Here, the subtyping module was evaluated using 795 clinical Klebsiella isolates, with whole genome sequences used to assess for blaKPC and p019. For the isolates identified as KPC positive by sequencing, the overall sensitivity of the MALDI-ToF MS subtyping module was 239/574 (42%) with 100% specificity. For the isolates harboring p019, the subtyping module showed a sensitivity of 97% (239/246) and a specificity of 100%. The subtyping module had poor sensitivity for the detection of blaKPC-positive Klebsiella isolates, albeit exhibiting excellent specificity. The poor sensitivity was a result of p019 being present in only 43% of the blaKPC-positive Klebsiella isolates.