The Kocurious case of Noodlococcus: genomic insights into Kocuria rhizophila from characterisation of a laboratory contaminant

Abstract

The laboratory contaminant strain Noodlococcus was named for its coccoid cells and unusual colony morphology, which resembled a pile of noodles. Along with laboratory characterisation and electron microscopy, we generated a complete Noodlococcus genome sequence using Illumina and Oxford Nanopore data. The genome consisted of a single, circular, 2,732,108 bp chromosome that shared 97.5% average nucleotide identity (ANI) with the Kocuria rhizophila type strain TA68. We identified genomic features involved in replication (oriC), carotenoid synthesis (crt) and genome defence (CRISPR-Cas) and discovered four novel mobile elements (ISKrh4-7). Despite its environmental ubiquity and relevance to food production, bioremediation and human medicine, there have been few genomic studies of the Kocuria genus. We conducted a comparative, phylogenetic and pangenomic examination of all 257 publicly available Kocuria genomes, with a particular focus on the 56 that were identified as K. rhizophila. We found that there are two phylogenetically distinct clades of K. rhizophila, with within-clade ANI values of 96.7-100.0% and between-clade values of 89.5-90.4%. The second clade, which we refer to as Kocuria pseudorhizophila, exhibited ANI values of <95% relative to TA68 and should constitute a separate species. Delineation of the two clades would be consistent with the rest of the genus, where all other species satisfy the 95% ANI threshold criteria. Differences in the K. rhizophila and K. pseudorhizophila pangenomes likely reflect phenotypic as well as evolutionary divergence. This distinction is relevant to clinical and industrial settings, as strains and genomes from both clades are currently used interchangeably, which may lead to reproducibility issues and phenotype-genotype discordance. Investigating an innocuous laboratory contaminant has therefore provided useful insights into the understudied species K. rhizophila, prompting an unexpected reassessment of its taxonomy.

Keywords: Kocuria rhizophila; contaminant; genomics; pangenome; taxonomy.

Fig. 1.. Morphology of laboratory contaminant Noodlococcus. (a, b) Photographs of the original Noodlococcus colony that was found growing on BHI agar. The smaller, white colonies were E. faecium that had re-grown after being picked for further culturing 11 days prior. (c) Phase-contrast micrograph of Gram-stained Noodlococcus. (d–f) Scanning electron micrographs of Noodlococcus cells.

Fig. 2.. Genomic features of Noodlococcus. All parts drawn to the same scale from GenBank accession CP097204. The extents and orientations of genes are indicated by labelled arrows beneath the horizontal lines that represent segments of the Noodlococcus genome. (a) Origin of replication. The oriC region is magnified 4.5× above to display fine-scale features as indicated in the key to the right. (b) Carotenoid synthesis region. The extents of regions that determine carotenoid synthesis and a putative ABC transporter are marked by labelled lines above. (c) CRISPR-Cas locus. The extent of the CRISPR locus is marked above. Each short, vertical pink line represents a copy of the 28 bp repeat unit, the sequence of which is shown below. (d) IS. Coloured vertical lines represent novel IS found in the Noodlococcus chromosome, grouped according to family membership. Terminal inverted repeats are shown as short black arrows and a seekRNA-determining region as a small black rectangle. (e) IS characteristics. Table outlining the features of novel IS characterised here. *, size estimate that requires experimental validation (see text); IR, terminal inverted repeats (identical bases/total bases).

Fig. 3.. Core genome phylogeny and pairwise ANIs of K. rhizophila genomes (n=51). Maximum likelihood phylogeny was inferred with the GTR+F+I+R5 substitution model using IQTREE. Ultrafast bootstrap supports are labelled on each branch, and the scale bar represents substitutions per site. K. tytonicola strains 473 and DSM 104133 were used as outgroups. Pairwise ANIs were calculated with FastANI and visualised with pheatmap. Proposed species demarcation is indicated by the group labels on the right side of the figure. Presence/absence of CRISPR (Cas genes), prophages and IS are shown to the right of the tree. Type strain TA68 is highlighted with a brown circle, and Noodlococcus is highlighted with a yellow circle.

Fig. 4.. ANIs between various Kocuria taxa. K. rhizophila (n=51), true rhizophila (n=35) and pseudorhizophila (n=16). All non-rhizophila Kocuria spp. only include genomes of species groups that contained >2 genomes (n=141).

Fig. 5.. Population-structure-aware pangenome of K. rhizophila. Number of gene clusters of the K. rhizophila pangenome (both clades) from each distribution class. Clade-specific gene clusters exclusively present in either rhizophila (yellow) or pseudorhizophila (purple) are shown for each distribution class. Core: >95% presence; shell: <95% and >15% presence; cloud: <15% presence. Distribution class definitions taken from [68].