Development of the cutaneous microbiome in the preterm infant: A prospective longitudinal study

Abstract

Background: Neonatal sepsis in preterm infants is often due to organisms that colonize the skin including Staphylococcus spp. and Candida spp. Development and maturation of the skin microbiome in the neonatal period, especially in preterm infants, may be critical in preventing colonization with pathogens and subsequent progression to neonatal sepsis. Development of the skin microbiome in preterm infants or its determinants in the first 4 weeks of life has not been evaluated.

Methods: We evaluated the skin microbiome from three body sites, antecubital fossa, forehead and gluteal region, in a prospective cohort of 15 preterm (birth weight < 1500 g and < 32 weeks of gestation) and 15 term neonates. The microbiome community membership and relative abundance were evaluated by amplification and sequencing the bacterial V3-V5 region of the16S rRNA gene on the 454 GS FLX platform. We used linear mixed effects models to analyze longitudinal data.

Results: The structure and composition of the skin microbiome did not differ between the three sampling sites for term and preterm infants in the neonatal period. However, skin bacterial richness was positively associated with gestational age in the first four weeks of life. Intravenous antibiotics negatively impacted the bacterial diversity of the skin but we did not see differences with respect to feeding or mode of delivery.

Conclusions: Gestational age, which influences the maturity of skin structure and function, is associated with the development of the preterm cutaneous microbiome. Understanding the maturation of a healthy skin microbiome, prevention of pathogen colonization and its role in the development of immunity will be pivotal in the development of novel interventions to prevent infections in critically ill preterm infants.

Fig 1. Longitudinal and site-specific development of the cutaneous microbiome in preterm and term infants.

Observed Operational Taxonomic Units (OTUs) (mean ± SEM) and Shannon diversity index (mean ± SEM) for preterm (Fig 1A and 1B) and term infants (Fig 1C and 1D) are plotted by sampling week and skin site. Term infants were sampled at 2 time points (birth and between 2–3 weeks of age) and preterm infants every week from birth for 5 time-points. For each skin site, the number of OTUs and Shannon diversity index in preterm infants decreased between week 1 and week 2, and then increased over time until week 5. The trend for diversity over time was not significantly different between the three sites in our linear mixed effects model for observed OTUs and Shannon diversity index in preterm infants. Overall, bacterial diversity tended to be higher for term infants compared to preterm infants (p = 0.04 for main effect of term vs. preterm status for Shannon diversity index; p = 0.05 for observed number of OTUs).

Fig 2. Hierarchical clustering for preterm and term infants, based on relative abundance at the genus level.

Each column represents a sample, with darker red indicating higher relative abundance. The top 10 most abundant genera are shown for preterm (2A) and term (2B) infants, separately. The top four rows show metadata for each sample namely use of significant antibiotics, mode of delivery, week of sample collection and sampling site.

Fig 3. Gestational age is associated with the skin microbiome (at baseline) in preterm infants.

3A. Average Observed OTU: Each dot represents the average number of Operational Taxonomic Units (OTUs) for the subject, across the sites. The average OTU is linearly regressed against gestational age at baseline samples and shaded areas are the 95% confidence intervals for the regression line. Gestational age is significantly associated with average number of OTUs, when both preterm and term infants are considered together (R = 0.39, p = 0.04). 3B. Average Shannon diversity index: Each dot represents the average Shannon diversity index for the subject, across the sites. The average Shannon diversity index is linearly regressed against gestational age at baseline samples and shaded areas are the 95% confidence intervals for the regression line. Gestational age is significantly associated with average Shannon diversity index, when both preterm and term infants are considered together (R = 0.49, p < 0.009).

Fig 4. The effects of antibiotic exposure on the development of the skin microbiome.

Alpha diversity indices of operational taxonomic units (OTUs) (mean ±SEM) are depicted in panels A, B and C for antecubital fossa, forehead and gluteal region respectively and Shannon diversity index (mean ±SEM) in panels D, E and F for antecubital fossa, forehead and gluteal region respectively in preterm neonates who were exposed to significant antibiotics (> 48 hr) compared to those not exposed. Exposure to significant antibiotics decreases alpha diversity measures in the preterm skin microbiome in the neonatal period.