A comprehensive phenotypic and comparative genomic investigation of bacterial strain Mich-4, originally isolated from the upper respiratory tract of a patient in Michigan, has established its precise identity as Elizabethkingia bruuniana while delineating its extensive multidrug-resistant resistome and virulence repertoire. Routine clinical diagnostic protocols frequently struggle to distinguish closely related species within the genus Elizabethkingia, and matrix-assisted laser desorption/ionization time-of-flight mass spectrometry initially misidentified the clinical isolate as Elizabethkingia meningoseptica. By deploying whole-genome sequencing alongside digital DNA-DNA hybridization, average nucleotide identity calculations, and phylogenomic reconstructions across a comparative panel of 47 Elizabethkingia genomes, researchers resolved the taxonomic misclassification and illuminated the evolutionary architecture of this emerging opportunistic pathogen, which has increasingly been implicated in severe human conditions such as bacteremia, septic shock, and postoperative endophthalmitis.

The genus Elizabethkingia comprises aerobic, nonmotile, Gram-negative bacilli that naturally inhabit water, soil, plants, and insects, but have gained prominence as opportunistic nosocomial pathogens capable of causing life-threatening infections in hospitalized, elderly, and immunocompromised patients. While species such as Elizabethkingia anophelis, Elizabethkingia meningoseptica, and Elizabethkingia miricola have been widely documented, newer taxa including Elizabethkingia bruuniana, Elizabethkingia ursingii, and Elizabethkingia occulta remain poorly characterized. In clinical diagnostic laboratories, overlapping biochemical phenotypes and incomplete commercial reference databases routinely obscure these emerging organisms, often grouping them generically under unclassified Elizabethkingia designations. The misidentification of strain Mich-4 mirrors previous historical revisions where isolates such as the termite gut symbiont BM10 and human clinical isolate EM798-26 were originally cataloged as Elizabethkingia miricola before whole-genome assessments reassigned them to Elizabethkingia bruuniana, demonstrating the critical diagnostic necessity of high-resolution genomic approaches.

Physiological and morphological characterization of strain Mich-4, which was recovered from a respiratory specimen on February 6, 2016, revealed straight, rod-shaped bacterial cells measuring approximately 1.6 micrometers in length by 0.5 micrometers in width with well-defined cell boundaries and smooth outer surfaces under negative-stain electron microscopy. The isolate stained Gram-negative, exhibited negative motility, and yielded positive reactions for oxidase, catalase, and indole production. Cultured aerobically on 5 percent sheep blood agar at 35 degrees Celsius for 24 hours, colonies displayed smooth, gray-white morphology with gamma-hemolysis, leaving surrounding red blood cells intact without discoloration or clearing. Growth on tryptic soy agar produced smooth gray colonies that lacked the characteristic yellow pigmentation frequently observed in other Elizabethkingia lineages. Comprehensive metabolic profiling using the Biolog GEN III MicroPlate system showed that Mich-4 actively metabolized an extensive array of carbohydrates and amino acids, including D-glucose, D-mannose, D-fructose, D-trehalose, D-maltose, gentiobiose, dextrin, N-acetyl-D-glucosamine, L-alanine, L-histidine, and L-glutamic acid, while exhibiting physiological sensitivity to elevated salinity at 4 percent sodium chloride, surfactant exposure from Niaproof 4, and acidic conditions below pH 5.0.

High-throughput genomic sequencing of Mich-4 was conducted on an Illumina MiSeq platform utilizing paired-end 2 by 250 base pair chemistry with a 500-cycle reagent kit, yielding 1,470,528 raw read pairs totaling 735,264,000 raw bases. Sequencing quality metrics indicated that 82.70 percent of raw bases achieved Phred scores of at least Q20 and 78.97 percent met or exceeded Q30, with forward reads reaching 87.23 percent Q30 and reverse reads reaching 70.71 percent Q30. De novo assembly performed via CLC Genomics Workbench generated a 4,201,659 base pair draft genome structured across 22 contigs with an N50 of 434,357 base pairs, an estimated raw sequencing depth of approximately 175-fold coverage, and an overall guanine-cytosine content of 35.80 percent. Assembly benchmarking via QUAST and BUSCO confirmed excellent genomic integrity, recording 99.1 percent complete BUSCO genes with 0.9 percent fragmented and zero missing markers, while CheckM analysis calculated 100.0 percent completeness, 1.08 percent contamination, and zero strain heterogeneity. Independent annotation pipelines further demonstrated concordance, with the NCBI Prokaryotic Genome Annotation Pipeline predicting 3,806 protein-coding genes, 12 pseudogenes, 42 tRNAs, four rRNAs, and three non-coding RNAs, closely matching Prokka annotations of 3,851 coding sequences, 41 tRNAs, and four rRNAs.

To rigorously verify the taxonomic standing of strain Mich-4, investigators implemented standardized phylogenomic and overall genome relatedness indices against reference genomes across the genus Elizabethkingia. Taxonomic assignment via the Genome Taxonomy Database Toolkit classified Mich-4 firmly within Elizabethkingia bruuniana, demonstrating an average nucleotide identity of 98.18 percent to reference strain G0146. Pairwise average nucleotide identity computed using FastANI showed 98.10 percent identity between Mich-4 and the type strain Elizabethkingia bruuniana CIP111191T, easily surpassing the accepted 95 percent species boundary, while identity scores between Mich-4 and other Elizabethkingia species fell between 81.6 percent and 93.5 percent. Digital DNA-DNA hybridization calculated via the Genome-to-Genome Distance Calculator against the type strain yielded values of 88.3 percent under formula 1 with a 95 percent confidence interval of 84.8 to 91.0 percent, 81.3 percent under formula 2 with a confidence interval of 78.4 to 83.9 percent, and 89.9 percent under formula 3 with a confidence interval of 87.1 to 92.1 percent. The corresponding statistical probabilities of digital DNA-DNA hybridization reaching or exceeding the 70 percent threshold were 96.61 percent, 91.52 percent, and 99.40 percent, respectively, confirming species assignment beyond doubt.

Comparative pangenome analysis evaluated across 12 Elizabethkingia bruuniana genomes revealed an open pangenome structure comprising 6,301 total gene clusters, separated into 3,062 core genes present across all 12 strains, 353 soft-core genes found in 11 genomes, 1,295 shell genes distributed across two to 10 genomes, and 1,591 cloud genes unique to single genomes. Modeling the pangenome accumulation trajectory using Heaps law yielded an equation of y equals 3715.522 multiplied by x raised to the power of 0.329 with a coefficient of determination of 0.9995. The positive exponent of 0.329 indicates that the gene repertoire of Elizabethkingia bruuniana continues to expand as additional isolates are sampled, confirming substantial genomic plasticity. Conversely, core genome decay modeling decayed according to an exponential curve approaching a theoretical asymptote of approximately 1,935 genes with an R-squared value of 0.9285. A focused UpSet intersection analysis among five closely related strains within the Mich-4 clade identified 3,722 shared genes, 228 accessory genes shared by subsets, and 417 strain-specific genes, with strain CBZ possessing the highest number of unique genes at 197, illustrating notable intraspecies divergence.

Functional enrichment assessments comparing accessory and core gene components using Clusters of Orthologous Groups annotations revealed significant functional partitioning within Elizabethkingia bruuniana. Following Benjamini-Hochberg adjustment, nine functional categories were significantly overrepresented in the accessory genome at a false discovery rate below 0.05. Mobilome elements representing prophages and transposons exhibited the most pronounced overrepresentation, with an odds ratio of 15.90 and a false discovery rate of 8.50 times 10 to the negative 13th power. Other significantly enriched accessory categories included inorganic ion transport and metabolism with an odds ratio of 3.32, defense mechanisms with an odds ratio of 2.88, cell wall, membrane, and envelope biogenesis with an odds ratio of 2.28, and signal transduction mechanisms with an odds ratio of 2.17. Further evaluation of 142 rare accessory gene clusters in Mich-4—defined as appearing in Mich-4 and no more than five of the 46 comparative genomes—identified specialized determinants for oxidative stress response and redox homeostasis including thioredoxin trxA, methionine sulfoxide reductases msrA and msrB, thiol-disulfide interchange protein dsbD, iron acquisition factor fecA, cell-envelope modifying enzymes arnC and mshA, transcriptional regulators exsA, rhaS, and rhaR, and TIR-domain cyclic ADP-ribose synthases thsB1 and thsB2.

In vitro antimicrobial susceptibility testing conducted using the automated VITEK 2 system demonstrated an extensive multidrug resistance profile in strain Mich-4. The isolate exhibited complete phenotypic resistance to all evaluated beta-lactam antibiotics, recording minimum inhibitory concentrations of 16 micrograms per milliliter or greater for meropenem, 64 micrograms per milliliter or greater for cefazolin, ceftriaxone, and aztreonam, 32 micrograms per milliliter or greater for cefotaxime, ampicillin, and ampicillin/sulbactam, 64 micrograms per milliliter or greater for piperacillin, and 128 micrograms per milliliter or greater for piperacillin/tazobactam. Mich-4 also displayed resistance to aminoglycosides, yielding minimum inhibitory concentrations of 64 micrograms per milliliter or greater for amikacin and 16 micrograms per milliliter or greater for both gentamicin and tobramycin, alongside resistance to nitrofurantoin at 128 micrograms per milliliter. Conversely, the strain demonstrated preserved susceptibility to trimethoprim-sulfamethoxazole with a minimum inhibitory concentration of 40 micrograms per milliliter and ciprofloxacin at 1 microgram per milliliter, while showing intermediate susceptibility to tigecycline at 4 micrograms per milliliter, contrasting sharply with previously documented Taiwanese isolates that were uniformly resistant to trimethoprim-sulfamethoxazole and largely susceptible to tigecycline.

Genomic resistome profiling through the Comprehensive Antibiotic Resistance Database and AMRFinderPlus uncovered the underlying molecular determinants responsible for this broad antimicrobial resistance phenotype. The beta-lactam resistance observed in Mich-4 was directly explained by the presence of three conserved, chromosomally encoded beta-lactamase gene families: blaB and blaGOB, encoding Ambler class B metallo-beta-lactamases capable of hydrolyzing carbapenems and cephalosporins while resisting clinical beta-lactamase inhibitors, and blaCME, encoding an Ambler class C extended-spectrum cephalosporinase. Aminoglycoside resistance was genetically supported by the presence of aadS, encoding an aminoglycoside adenylyltransferase conferring enzymatic drug inactivation, as well as ranA and ranB, which encode an ATP-binding cassette multidrug efflux transporter system. However, observed resistance to nitrofurantoin and intermediate susceptibility to tigecycline, along with susceptibility to ciprofloxacin and trimethoprim-sulfamethoxazole, were not linked to specific annotated resistance genes in the databases, illustrating that phenotypic resistance profiles can involve uncharacterized chromosomal mutations, altered membrane permeability, or differential regulatory pathways not captured by current reference repositories.

Screening against the Virulence Factor Database established that Elizabethkingia bruuniana possesses an extensive and conserved pathogenic armamentarium across both core and accessory genomic compartments. Ubiquitous factors included oxidative stress mitigation genes katA and katG, molecular chaperones groEL and htpB, metabolic adaptation enzymes icl and pgi, translation factors tuf and tufA, and urease operon components ureA and ureB. Urease enzymes hydrolyze urea to generate carbon dioxide and ammonia, thereby elevating ambient pH and facilitating persistent colonization within hostile mucosal microenvironments, a mechanism well recognized in respiratory pathogens such as Haemophilus influenzae. Notably, while the urease operon was conserved in Mich-4 and other Elizabethkingia bruuniana strains, it was completely absent from analyzed genomes of Elizabethkingia anophelis and Elizabethkingia meningoseptica, suggesting lineage-specific ecological adaptation. Furthermore, while core lipopolysaccharide biosynthesis enzymes like lpxD, rfbA, rmlA, and wbtL were uniformly maintained across all 12 Elizabethkingia bruuniana genomes, the GDP-mannose 4,6-dehydratase gene gmd was absent in Mich-4 and six related strains, highlighting subtle structural variations in surface antigenicity.

A major structural discovery in the study was the characterization of a Type VI Secretion System subtype III locus, designated T6SSiii, representing the first formal identification of this contractile macromolecular nanomachine in Elizabethkingia bruuniana. Encoded across 11 of the 12 analyzed genomes including Mich-4, the complete cluster incorporates contractile sheath components TssB and TssC, inner tube protein TssD, spike protein TssI, ATPase motor TssH, baseplate proteins TssE, TssF, TssG, and TssK, and five Bacteroidota-specific membrane-complex subunits TssN, TssO, TssP, TssQ, and TssR. Phylogenetic trees constructed from concatenated T6SSiii structural proteins exhibited topological congruence with the core-genome phylogeny, showing identical branching patterns without evidence of long-branch displacement or clade shuffling, which demonstrates that the T6SSiii locus co-diverged with the core genome rather than arising from recent horizontal acquisition. Functioning as a molecular contact-dependent syringe, homologous T6SSiii systems in related Bacteroidota such as Flavobacterium johnsoniae and Bacteroides fragilis deploy bactericidal effectors to kill bacterial rivals and secure colonization niches, indicating a comparable competitive role in Elizabethkingia biology.

The study also identified widespread ICEEa-like Type B conjugative transfer modules across 11 of the 12 Elizabethkingia bruuniana genomes, confirming the central contribution of integrative and conjugative elements to genomic diversification within the species. The detected modules encompassed core mobilization and mating-pair apparatus genes including relaxase, type IV coupling protein VirD4, VirB4/TraG, and conjugative transfer proteins TraE through TraQ. Marker-protein sequence alignments revealed that most Elizabethkingia bruuniana elements possess a mosaic structure, exhibiting mixed homology where relaxase and coupling proteins align most closely with the ICEEaI reference element of Elizabethkingia anophelis, while VirB4/TraG and TraJ align with ICEEaII. Strain BM10 displayed homology to the ICEEaIII reference, whereas strain SBRL-21-126 lacked detectable conjugative elements altogether. The authors noted that predicted virulence determinants predominantly resided outside these mobile genomic islands, suggesting that intrinsic chromosomal loci rather than recently acquired mobile elements form the primary foundation of virulence in strain Mich-4.

The authors contextualized their findings by emphasizing several critical analytical constraints and avenues for ongoing investigation. Because short-read sequencing assemblies cannot fully resolve repetitive sequence tracts, long-read sequencing approaches will be necessary to definitively map the exact integration sites, attachment sequences, and structural boundaries of these integrative mobile elements. Additionally, in silico identification of resistance genes and secretion clusters establishes genetic presence but does not confirm transcriptional activity or protein expression, necessitating targeted functional assays and transcriptomic evaluations. Furthermore, because the Clinical and Laboratory Standards Institute has not established species-specific breakpoints for Elizabethkingia, clinical interpretations rely on criteria defined for other non-Enterobacterales, a surrogate framework whose therapeutic predictive accuracy remains unvalidated. Expanding global genomic surveillance and functional research will be essential to track how this plastic, multidrug-resistant pathogen disseminates across healthcare institutions and natural environments.

Subject of Research: Comparative genomic, taxonomic, and resistome characterization of Elizabethkingia bruuniana strain Mich-4

Article Title: Elizabethkingia bruuniana strain Mich-4: molecular identification, genome analysis, resistome characterization, and virulence profiling

Article References: Omburo, B., Riegler, M., Walker, E. D., & Chen, S. (2026). Elizabethkingia bruuniana strain Mich-4: molecular identification, genome analysis, resistome characterization, and virulence profiling. Frontiers in Cellular and Infection Microbiology, 16, Article 1941985. https://doi.org/10.3389/fcimb.2026.1941985

Image Credits: AI Generated

DOI: 10.3389/fcimb.2026.1941985

Keywords: Elizabethkingia bruuniana, Mich-4, comparative genomics, pangenome analysis, antimicrobial resistance, beta-lactamase, multidrug resistance, MALDI-TOF MS, digital DNA-DNA hybridization, average nucleotide identity, Type VI secretion system, T6SSiii, integrative and conjugative elements, ICEEa-like elements, urease operon, virulence profiling, opportunistic pathogen, respiratory infection