Robyn Baker

Microbes in the neonatal intensive care unit resemble those found in the gut of premature infants

BackgroundThe source inoculum of gastrointestinal tract (GIT) microbes is largely influenced by delivery mode in full-term infants, but these influences may be decoupled in very low birth weight (VLBW, <1,500 g) neonates via conventional broad-spectrum antibiotic treatment. We hypothesize the built environment (BE), specifically room surfaces frequently touched by humans, is a predominant source of colonizing microbes in the gut of premature VLBW infants. Here, we present the first matched fecal-BE time series analysis of two preterm VLBW neonates housed in a neonatal intensive care unit (NICU) over the first month of life.ResultsFresh fecal samples were collected every 3 days and metagenomes sequenced on an Illumina HiSeq2000 device. For each fecal sample, approximately 33 swabs were collected from each NICU room from 6 specified areas: sink, feeding and intubation tubing, hands of healthcare providers and parents, general surfaces, and nurse station electronics (keyboard, mouse, and cell phone). Swabs were processed using a recently developed ‘expectation maximization iterative reconstruction of genes from the environment’ (EMIRGE) amplicon pipeline in which full-length 16S rRNA amplicons were sheared and sequenced using an Illumina platform, and short reads reassembled into full-length genes. Over 24,000 full-length 16S rRNA sequences were produced, generating an average of approximately 12,000 operational taxonomic units (OTUs) (clustered at 97% nucleotide identity) per room-infant pair. Dominant gut taxa, including Staphylococcus epidermidis, Klebsiella pneumoniae, Bacteroides fragilis, and Escherichia coli, were widely distributed throughout the room environment with many gut colonizers detected in more than half of samples. Reconstructed genomes from infant gut colonizers revealed a suite of genes that confer resistance to antibiotics (for example, tetracycline, fluoroquinolone, and aminoglycoside) and sterilizing agents, which likely offer a competitive advantage in the NICU environment.ConclusionsWe have developed a high-throughput culture-independent approach that integrates room surveys based on full-length 16S rRNA gene sequences with metagenomic analysis of fecal samples collected from infants in the room. The approach enabled identification of discrete ICU reservoirs of microbes that also colonized the infant gut and provided evidence for the presence of certain organisms in the room prior to their detection in the gut.

Variability in Cerebral Oxygen Delivery Is Reduced in Premature Neonates Exposed to Chorioamnionitis

Premature infants exposed to chorioamnionitis are at increased risk for periventricular leukomalacia (PVL) and intraventricular hemorrhage (IVH), lesions that may result from inflammation and/or fluctuations in cerebral blood flow. The effect of chorioamnionitis on near-infrared spectroscopy (NIRS) measures of cerebral oxygen delivery has not been evaluated previously. Forty-nine infants born at 25–31 6/7 wk gestation underwent NIRS examination on d 1, 2, 3, and 7 of life. Variability in NIRS tracings was analyzed by partitioning each tracing into three components: long-term, intermediate, and short-term variability; the latter two components were analyzed. Chorioamnionitis-exposed infants manifest reduced intermediate variability in cerebral oxygenated and deoxygenated Hb but not total Hb. Infants with severe IVH/PVL had the lowest intermediate variability on d 1. Short-term variability was similar between chorioamnionitis-exposed and unexposed infants, and between infants with versus without severe IVH or PVL. We conclude that intermediate-term variability in NIRS cerebral oxygen delivery is reduced in chorioamnionitis-exposed infants. We speculate that intermediate variability represents the important time frame for evaluating the pathogenesis of perinatal brain injury. Further studies are needed to determine how these findings relate to cerebral blood flow autoregulation and oxygen utilization in premature infants.

Gut bacteria are rarely shared by co-hospitalized premature infants,regardless of necrotizing enterocolitis development

Premature infants are highly vulnerable to aberrant gastrointestinal tract colonization, a process that may lead to diseases like necrotizing enterocolitis. Thus, spread of potential pathogens among hospitalized infants is of great concern. Here, we reconstructed hundreds of high-quality genomes of microorganisms that colonized co-hospitalized premature infants, assessed their metabolic potential, and tracked them over time to evaluate bacterial strain dispersal among infants. We compared microbial communities in infants who did and did not develop necrotizing enterocolitis. Surprisingly, while potentially pathogenic bacteria of the same species colonized many infants, our genome-resolved analysis revealed that strains colonizing each baby were typically distinct. In particular, no strain was common to all infants who developed necrotizing enterocolitis. The paucity of shared gut colonizers suggests the existence of significant barriers to the spread of bacteria among infants. Importantly, we demonstrate that strain-resolved comprehensive community analysis can be accomplished on potentially medically relevant time scales. DOI: http://dx.doi.org/10.7554/eLife.05477.001

Low blood pressure among very-low-birth-weight infants with fetal vessel inflammation

OBJECTIVE: To test the hypothesis that fetal vessel inflammation (FVI: funisitis and/or fetal vasculitis) is associated with lower blood pressure (BP) over the first week of life and an increased risk of periventricular leukomalacia (PVL) among premature infants.STUDY DESIGN: A total of 255 infants born at <1350 g to normotensive mothers were stratified by gestational age (GA) and grouped by presence/absence of FVI on placental pathology. Daily highest (Hi) and lowest (Lo) systolic BP (BPsys), mean BP (BPmn) and diastolic BP (BPdia) over first 7 days of life were analyzed by repeated measures ANOVA and regression analysis. Cranial ultrasounds were obtained at 2 weeks of life.RESULTS: Infants ≥30 weeks gestation with FVI had lower HiBPsys, HiBPmn, HiBPdia, LoBPsys, LoBPmn and LoBPdia (p<0.001) than did infants without FVI. Infants with PVL (all ≤27 weeks gestation) had lower LoBPmn and LoBPdia (p<0.01) than controls. FVI did not increase the risk of PVL in these infants.CONCLUSION: FVI and PVL are associated with reduced BP over the first week of life.

Mucosa-associated bacterial diversity in necrotizing enterocolitis.

Background Previous studies of infant fecal samples have failed to clarify the role of gut bacteria in the pathogenesis of NEC. We sought to characterize bacterial communities within intestinal tissue resected from infants with and without NEC. Methods 26 intestinal samples were resected from 19 infants, including 16 NEC samples and 10 non-NEC samples. Bacterial 16S rRNA gene sequences were amplified and sequenced. Analysis allowed for taxonomic identification, and quantitative PCR was used to quantify the bacterial load within samples. Results NEC samples generally contained an increased total burden of bacteria. NEC and non-NEC sample sets were both marked by high inter-individual variability and an abundance of opportunistic pathogens. There was no statistically significant distinction between the composition of NEC and non-NEC microbial communities. K-means clustering enabled us to identify several stable clusters, including clusters of NEC and midgut volvulus samples enriched with Clostridium and Bacteroides. Another cluster containing both NEC and non-NEC samples was marked by an abundance of Enterobacteriaceae and decreased diversity among NEC samples. Conclusions The results indicate that NEC is a disease without a uniform pattern of microbial colonization, but that NEC is associated with an abundance of strict anaerobes and a decrease in community diversity.

Prophylactic Indomethacin Reduces Grades III and IV Intraventricular Hemorrhages when Compared to Early Indomethacin Treatment of a Patent Ductus Arteriosus

OBJECTIVE: To determine the relative risk of severe intraventricular hemorrhage (IVH) between two very early indomethacin treatment strategies.STUDY DESIGN: Retrospective chart review of infants <29 weeks gestation and <1350 g who received either indomethacin prophylaxis or very early echocardiography with indomethacin treatment only if the ductus arteriosus was patent.RESULTS: A total of one hundred and two infants received prophylactic indomethacin (pINDO). Echochardiography was performed on 158 infants, of whom 117 received indomethacin. Infants receiving pINDO had lower gestational age, but similar birth weight, gender, race, antenatal steroid exposure, delivery mode, Apgar scores, and need for resuscitation as infants evaluated by echocardiography. Grades III to IV IVH was observed less frequently in infants who received pINDO (OR 0.27, 95% CI 0.10 to 0.77, p=0.014). Frequency of side effects and recurrent patent ductus arteriosus did not differ between treatment groups.CONCLUSION: pINDO reduces severe IVH when compared to an early echocardiography strategy.

Identical bacterial populations colonize premature infant gut, skin, and oral microbiomes and exhibit different in situ growth rates

The initial microbiome impacts the health and future development of premature infants. Methodological limitations have led to gaps in our understanding of the habitat range and subpopulation complexity of founding strains, as well as how different body sites support microbial growth. Here, we used metagenomics to reconstruct genomes of strains that colonized the skin, mouth, and gut of two hospitalized premature infants during the first month of life. Seven bacterial populations, considered to be identical given whole-genome average nucleotide identity of >99.9%, colonized multiple body sites, yet none were shared between infants. Gut-associated Citrobacter koseri genomes harbored 47 polymorphic sites that we used to define 10 subpopulations, one of which appeared in the gut after 1 wk but did not spread to other body sites. Differential genome coverage was used to measure bacterial population replication rates in situ. In all cases where the same bacterial population was detected in multiple body sites, replication rates were faster in mouth and skin compared to the gut. The ability of identical strains to colonize multiple body sites underscores the habit flexibility of initial colonists, whereas differences in microbial replication rates between body sites suggest differences in host control and/or resource availability. Population genomic analyses revealed microdiversity within bacterial populations, implying initial inoculation by multiple individual cells with distinct genotypes. Overall, however, the overlap of strains across body sites implies that the premature infant microbiome can exhibit very low microbial diversity.

Evidence for persistent and shared bacterial strains against a background of largely unique gut colonization in hospitalized premature infants

The potentially critical stage of initial gut colonization in premature infants occurs in the hospital environment, where infants are exposed to a variety of hospital-associated bacteria. Because few studies of microbial communities are strain-resolved, we know little about the extent to which specific strains persist in the hospital environment and disperse among infants. To study this, we compared 304 near-complete genomes reconstructed from fecal samples of 21 infants hospitalized in the same intensive care unit in two cohorts, over 3 years apart. The genomes represent 159 distinct bacterial strains, only 14 of which occurred in multiple infants. Enterococcus faecalis and Staphylococcus epidermidis, common infant gut colonists, exhibit diversity comparable to that of reference strains, inline with introduction of strains from infant-specific sources rather than a hospital strain pool. Unlike other infants, a pair of sibling infants shared multiple strains, even after extensive antibiotic administration, suggesting overlapping strain-sources and/or genetic selection drive microbiota similarities. Interestingly, however, five strains were detected in infants hospitalized three years apart. Three of these were also detected in multiple infants in the same year. This finding of a few widely dispersed and persistent bacterial colonizers despite overall low potential for strain dispersal among infants has implications for understanding and directing healthy colonization.

Concentrations and Sources of Airborne Particles in a Neonatal Intensive Care Unit

Premature infants in neonatal intensive care units (NICUs) have underdeveloped immune systems, making them susceptible to adverse health consequences from air pollutant exposure. Little is known about the sources of indoor airborne particles that contribute to the exposure of premature infants in the NICU environment. In this study, we monitored the spatial and temporal variations of airborne particulate matter concentrations along with other indoor environmental parameters and human occupancy. The experiments were conducted over one year in a private-style NICU. The NICU was served by a central heating, ventilation and air-conditioning (HVAC) system equipped with an economizer and a high-efficiency particle filtration system. The following parameters were measured continuously during weekdays with 1-min resolution: particles larger than 0.3 μm resolved into 6 size groups, CO2 level, dry-bulb temperature and relative humidity, and presence or absence of occupants. Altogether, over sixteen periods of a few weeks each, measurements were conducted in rooms occupied with premature infants. In parallel, a second monitoring station was operated in a nearby hallway or at the local nurses’ station. The monitoring data suggest a strong link between indoor particle concentrations and human occupancy. Detected particle peaks from occupancy were clearly discernible among larger particles and imperceptible for submicron (0.3–1 μm) particles. The mean indoor particle mass concentrations averaged across the size range 0.3–10 μm during occupied periods was 1.9 μg/m3, approximately 2.5 times the concentration during unoccupied periods (0.8 μg/m3). Contributions of within-room emissions to total PM10 mass in the baby rooms averaged 37–81%. Near-room indoor emissions and outdoor sources contributed 18–59% and 1–5%, respectively. Airborne particle levels in the size range 1–10 μm showed strong dependence on human activities, indicating the importance of indoor-generated particles for infant’s exposure to airborne particulate matter in the NICU.

Asymmetry of Cerebral Blood Flow Velocity in Low Birth Weight Infants

Background: In premature infants, intraventricular hemorrhage occurs more commonly in the left than the right hemisphere. We have demonstrated previously that cerebral blood flow velocity is lower in the left than the right middle cerebral artery in the first few hours after birth. This may be due to the open ductus arteriosus. Objective: To test the hypothesis that blood flow velocity is lower in the left than the right middle cerebral artery only when the ductus arteriosus is open. Study Design: Infants born at 25–33 weeks’ gestation were enrolled. Middle cerebral artery blood flow velocities and coefficients of variation were measured on the left, followed by the right, on days 1 and 7 of life. Echocardiography identified 67 infants (25–33 weeks, 517–2,371 g) whose ductus arteriosus was open on day 1 and closed on day 7. Results: Systolic (26.4 ± 7.4 vs. 29.6 ± 7.2 cm/s), mean (12.4 ± 4.0 vs. 15.6 ± 4.6 cm/s) and end-diastolic (5.3 ± 2.2 vs. 6.8 ± 2.9 cm/s) blood flow velocities were lower (p < 0.01) and the corresponding coefficients of variation were higher (p < 0.01) on the left on day 1. Neither the absolute Doppler blood flow velocities nor the coefficients of variation differed between the left and right sides on day 7. Conclusions: Blood flow velocity is lower and more variable in the left compared to the right middle cerebral artery on day 1 of life in premature infants. These differences are not found on day 7. We speculate that this difference is due to the associated ductus arteriosus patency.