Reducing viability bias in analysis of gut microbiota in preterm infants at risk of NEC and sepsis

Abstract

Necrotising enterocolitis (NEC) and sepsis are serious diseases of preterm infants that can result in feeding intolerance, the need for bowel resection, impaired physiological and neurological development, and high mortality rates. Neonatal healthcare improvements have allowed greater survival rates in preterm infants leading to increased numbers at risk of developing NEC and sepsis. Gut bacteria play a role in protection from or propensity to these conditions and have therefore, been studied extensively using targeted 16S rRNA gene sequencing methods. However, exact epidemiology of these conditions remain unknown and the role of the gut microbiota in NEC remains enigmatic. Many studies have confounding variables such as differing clinical intervention strategies or major methodological issues such as the inability of 16S rRNA gene sequencing methods to determine viable from non-viable taxa. Identification of viable community members is important to identify links between the microbiota and disease in the highly unstable preterm infant gut. This is especially important as remnant DNA is robust and persists in the sampling environment following cell death. Chelation of such DNA prevents downstream amplification and inclusion in microbiota characterisation. This study validates use of propidium monoazide (PMA), a DNA chelating agent that is excluded by an undamaged bacterial membrane, to reduce bias associated with 16S rRNA gene analysis of clinical stool samples. We aim to improve identification of the viable microbiota in order to increase the accuracy of clinical inferences made regarding the impact of the pretermgutmicrobiota on health and disease. Gutmicrobiota analysis was completed on stools from matched twins (n = 16) that received probiotics. Samples were treated with PMA, prior to bacterial DNA extraction. Meta-analysis highlighted a significant reduction in bacterial diversity in 68.8% of PMA treated samples as well as significantly reduced overall rare taxa abundance. Importantly, overall abundances of genera associated with protection from and propensity to NEC and sepsis such as: Bifidobacterium; Clostridium, and Staphylococcus sp. were significantly different following PMA-treatment. These results suggest non-viable cell exclusion by PMA-treatment reduces bias in gut microbiota analysis from which clinical inferences regarding patient susceptibility to NEC and sepsis are made.