Lactobacillus delbrueckii subsp. allosunkii and lactis as emerging human uropathogens in elderly patients

Abstract

Lactobacillus delbrueckii has been considered a very rare cause of human urinary tract infections (UTIs). However, little is known about its clinical significance and antimicrobial susceptibility, and genomic data from clinical isolates are lacking. This study aimed at analyzing clinical, microbiological, and genomic data of L. delbrueckii urinary isolates. All L. delbrueckii isolates collected from patients hospitalized in a French university hospital from 2014 to 2016 were included. Clinical and biological data were gathered. Species identification was performed by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry, and MICs were determined using the broth microdilution method. Whole genome sequencing (WGS) was conducted (Illumina MiSeq 2 × 300 bp), and genomes were compared using three approaches (multilocus sequence typing [MLST], average nucleotide identity [ANI], and core genome single nucleotide polymorphism [SNP]). From 2014 to 2016, 48 isolates of L. delbrueckii were recovered from the urine of 48 patients (mean age = 84 years; sex ratio M/F = 0.04). Nearly half (44%) of patients were diagnosed with a UTI, and all had significant cultures (≥10⁵ CFU/mL) with a positive direct examination in >90% of cases. The majority of isolates were susceptible to most antibiotics (especially β-lactams), whereas they seemed intrinsically resistant to fosfomycin and metronidazole. Subspecies identification was consistent across the three approaches, showing that most L. delbrueckii isolates belonged to subspecies allosunkii (n = 40; 83%), followed by subspecies lactis (n = 8; 17%). Two isolates were resistant to tetracycline (MIC >16 mg/L) and both harbored the tet(W) gene. This study demonstrates the uropathogenic role of L. delbruekii subspecies allosunkii and lactis, particularly in elderly female patients.IMPORTANCEThis largest case series of urinary tract infections (UTIs) caused by Lactobacillus delbrueckii clearly demonstrates the uropathogenic role of this species (especially the subspecies allosunkii) in human UTIs, particularly in elderly female patients and those with underlying comorbidities. This study may change practice in two ways: (i) clinical laboratories, which typically consider lactobacilli from urine samples as contaminants, may need to reassess this practice; (ii) patient care can be improved by prescribing appropriate antibiotics for these underdiagnosed UTIs. L. delbrueckii should be considered an actual pathogen when it is significantly found in the urine of predisposed patients with clinical and/or biological signs of infection. While matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry allows reliable identification of L. delbrueckii, there is also a need for better discrimination between subspecies (especially allosunkii and lactis). Since L. delbrueckii isolates are usually susceptible to many antibiotics, we recommend β-lactams (especially aminopenicillins) for the treatment of those UTIs.

Keywords: Lactobacillus delbrueckii; UTI; lactobacilli; subspecies; urine.

Fig 1

Minimum spanning tree based on MLST data from 568 L. delbrueckii isolates. The collection of isolates includes 480 isolates from NCBI, 40 strains described by Tanigawa et al. (Table S2) (31), and the 48 clinical isolates of the study. The seven housekeeping genes (fusA, gyrB, hsp60, ileS, pyrG, recA, and recG) were aligned against all genomes, and the tree was built from concatenated sequences using GrapeTree (https://github.com/achtman-lab/GrapeTree). Clinical isolates of the study are colored in yellow.

Fig 2

ANI analysis of the 48 isolates compared with L. delbrueckii reference subspecies genomes. The maximum-likelihood phylogenetic tree was constructed using a core genome SNP-based approach with IQ-TREE 2 (TVM+F+ASC+G4 best-fit model) and 1,000 ultrafast bootstrap iterations (http://www.iqtree.org/). The tree was visualized using iTOL (https://itol.embl.de). The reference strains included are L. delbrueckii subsp. allosunkii JCM 17838, L. delbrueckii subsp. lactis DSM 20072, L. delbrueckii subsp. bulgaricus ATCC 11842, L. delbrueckii subsp. delbrueckii DSM 20074, L. delbrueckii subsp. indicus JCM 15610, and L. delbrueckii subsp. jakobsenii DSM 26046.

Fig 3

Core genome SNP-based maximum-likelihood tree of 528 L. delbrueckii isolates. This data set comprises 480 isolates from NCBI and 48 clinical isolates from this study. The tree was constructed using IQ-TREE 2 (GTR+F+ASC+G4 best-fit model) with 1,000 ultrafast bootstrap iterations (http://www.iqtree.org/), based on 132,691 polymorphic sites within the core genome. Visualization was performed using iTOL (https://itol.embl.de). Clinical isolates from this study are indicated by black rectangles, and their names are highlighted in yellow.

Fig 4

Maximum-likelihood phylogenetic tree based on the alignment of 10,137 recombination-masked SNPs derived from 1,207 core gene sequences from the 40 L. delbrueckii subsp. allosunkii clinical isolates. A heatmap is provided to illustrate SNP distances among the isolates. The three primary lineages (1–3) and four clades (A–D) are indicated.