BACKGROUND: Urinary tract infections (UTIs) are the most common outpatient infections in the nation and are among the most serious bacterial infections encountered by pediatricians. UTIs have a recurrence rate of up to 50%, and in children, can result in life-long health consequences. Adult studies have begun to detect differences in the urobiomes of individuals with and without recurrent UTIs. Identified differences could lead to changes in prevention and treatment strategies as well as provide novel avenues for advanced diagnostics. With the latter in mind, we focused on analyses of voided urine samples that are expected to contain microbes from several adjacent niches (i.e. bladder, skin, genitals), and thus have a broader representation of microbes that are known to affect bladder colonization with commensal or pathogenic bacteria and provide relevant information from the most clinically available samples. As comparative studies in children are yet to be conducted, in this study, we test the feasibility of biomass analysis of voided urine using two different sequencing methods and samples from small cohorts of children without (noUTI) and with recurrent UTI (rUTI).
METHODS: 10 voided urine samples were collected in children without a history of UTI (noUTI) and 10 in children with a history of rUTI (at least 3 UTIs in the last 12 months) while recording demographic and clinical variables. Urine specimens were filtered through 0.22 μm membrane to collect biomass, and total DNA was extracted and sequenced using traditional short V4 amplicon sequencing using Illumina platform and LoopSeq synthetic long-read sequencing of the entire 16s rRNA gene.
RESULTS: All 20 samples produced V4 data, and 14/20 samples produced long-read data. Overall, data produced with both methods yielded similar results in terms of the types of bacteria present. Importantly, the long-read data was able to identify more reads at the species-level. The bacteria Peptoniphilus harei, and Anaerrococcus lactolytic, and Prevotella timonensis were found in the majority of samples (14/14, 12/14, 12/14 respectively). In comparing bacteria across the noUTI and rUTI cohorts, a few differences were identified in the rUTI cohort, including decreased Corynebacterium (median [IQR]) 3.65 [1.78, 4.88] noUTI; 0.20 [0.09, 0.25] rUTI; p = 0.085, and increased Finegoldia (median [IQR]) 0.56 [0.34, 2.01] noUTI; 2.89 [1.98, 4.39] rUTI, p = 0.064).
CONCLUSIONS: This pilot study demonstrates feasibility for evaluating the urobiome in children without and with rUTI and using newer sequencing technology that allows for species-level identification of bacteria. The composition of the pediatric urobiome was similar across sequencing methodology and consistent with our other studies using catheterized urine specimens. The finding that Corynebacterium is decreased in the rUTI population compared to controls coincides with our previously reported observation that Corynebacterium was decreased in children with 3 or more UTIs compared to those with a history of only 1 UTI. This study sets the stage for future detailed investigation into urogenital microbiome of healthy and rUTI pediatric patients with the goal of identifying novel diagnostic, prevention, and treatment strategies.
