Dynamics of faecal bacterial populations in early infancy as determined by massively parallel sequencing

Master Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title

University of Cape Town

Background: Meconium microbiota have recently gained great interest; however very few studies have included meconium specimens when longitudinally characterizing the infant GIT microbiota. This study therefore aimed to longitudinally characterize meconium microbiota profiles during the first seven months of life and to compare these profiles with those from maternal faecal specimens using quality controlled Illumina MiSeq sequencing data. Methods: We sampled infant meconium and maternal faecal specimens at birth, as well as two subsets of infant faecal specimens at 4-12 and 20-28 weeks of life. We extracted nucleic acid from faecal specimens using the automated QIAsymphony ® SP instrument. Using Illumina MiSeq technology, we sequenced the V4 region of the bacterial 16S rRNA gene. We determined whether sufficient reads were sequenced using accumulation curves; whether any contamination occurred; and whether our sequencing approach was reproducible. The relative abundances of taxonomically classified operational taxonomic units (OTUs), and the Shannon diversity and Bray Curtis dissimilarity indices served to characterize faecal specimens from participants. Log ratio biplots and generalized linear mixed models served to statistically determine differences between faecal bacterial profiles. Results: Faecal specimens were collected from 90 mothers and 107 infants at birth, 72 infants at 4-12 and 36 infants at 20 28 weeks of age. We classified OTUs from two non-template controls which were indicative of potential contamination. Correcting for contamination resulted in a loss of 10 % of OTUs classified. Our reproducibility analysis correlated with increased concentrations of template used during library preparation. Based on diversity measures, meconium specimens harboured the most diverse bacterial profiles. The highest proportions of OTUs classified from meconium belonged to the phylum Proteobacteria (60 %), while the phylum Firmicutes was most abundant at 4-12 weeks (49 %) and 20-28 weeks (64 %) of life. The phylum Actinobacteria was at its highest at 4-12 weeks of age (26 %) and its increased proportions were associated with breastfeeding at 6-10 weeks of life. Firmicutes constituted the majority (79 %) of bacteria from maternal faecal specimens. No mother- infant pairs clustered at any of the time points studied, but infant bacterial profiles became more adult-like with increased age. An increase in infant age significantly affected bacterial proportions of 87 OTUs. Interestingly, we observed that infants exposed to HIV had higher proportions of the genus Leuconostoc and higher diversity indices compared to HIV unexposed infants at 4-12 weeks of age. Conclusion: Our study highlights that reproducibility may be worsened by the use of low template concentrations during library preparation, which may also skew diversity measures. We conclude that meconium is not sterile and that infant faecal bacterial profiles become more adult-like with increased age.