Abstract
Preterm infants are often exposed to microbiota modifying factors in early life and carry an increased risk of respiratory disease beyond infancy. Long hospital stays, intensive care procedures and use of antibiotics are essential for the preterm infant’s survival but may leave lasting scars on their microbiota and their long-term health. Despite high prevalence of respiratory morbidities among preterm infants, the metagenome of respiratory microbiota has mostly remained uncharted. Methodological challenges of microbiota samples containing high host DNA and low microbial biomass have made the implementation of whole metagenomic sequencing for respiratory samples very demanding.
In this thesis, I first address the methodological challenges of upper respiratory samples analysis, using nasopharyngeal aspirates of very preterm infants. Our experimental set up compared several host DNA depletion and microbial DNA extraction methods to establish a standard operating protocol for metagenomic analysis. Using our protocol, we were able to achieve species and antibiotic resistance gene (ARG) resolution after whole metagenomic sequencing of both pooled and individual patient samples.
Further, we applied the protocol on 369 nasopharyngeal aspirate samples obtained from a cohort of 66 very preterm infants, sampled from birth until six months corrected age. Microbiome composition differed from that described in term infants. Interindividual variability, followed by postnatal age exerted the most significant impact on the overall microbiome composition. We observed a persistent effect of hospitalization until six months corrected age, and a transient effect of postnatal antibiotics that diminished until discharge from the hospital and was not visible at six months corrected age.
In the last part of the thesis, we investigated the nasopharyngeal antibiotic resistance profile in a sub-cohort of 36 infants, exposed to either ampicillin plus gentamycin for suspected early onset neonatal sepsis, or not receiving any postnatal antibiotics. The resistome composition strongly correlated with microbial communities. We found high-risk ARGs in preterm infants regardless of postnatal exposure and observed an increase in abundance and diversity of ARGs after discontinuation of antibiotics, which had a significant impact on the resistome composition but diminished until discharge. Additionally, we noticed a cumulative effect of pre- and postnatal antibiotics on the resistome of infants exposed to ampicillin plus gentamycin. Serratia species associated resistance profile showed persistence across all sampling time points and was still visible at six months corrected age.