Background & Aims

Dysbiosis of the intestinal microbiota has been associated with development of allergies in infants. However, it is not clear what microbes might contribute to this process. We investigated what microbe(s) might be involved in analyses of infant twins and mice.

Methods

We studied fecal specimens prospectively in a twin cohort (n = 30) and age-matched singletons (n = 14) born at National Taiwan University Children’s Hospital, Taipei, Taiwan, from April 2011 to March 2013. Clinical parameters (gestational age, birth body weight, mode of delivery and feeding, immunizations, and medical events) were recorded. Fecal samples were collected beginning immediately after birth and for 1 year; the children were followed until 3 years of age and allergic symptoms (repetitive and continuous for at least 6 months) were noted. A skin prick test was used to ascertain atopy. Bacterial communities in fecal samples were profiled by 16S ribosomal RNA-based polymerase chain reaction−temporal temperature gradient gel electrophoresis and next-generation sequencing. BALB/c mice without and with ovalbumin sensitization/challenge were infected with candidate bacteria by oral gauge intragastric intubation. Fecal, serum, lung, and colon tissue samples were collected from mice and analyzed for mechanisms of allergy development.

Results

During the investigation period, 20 children (45.5%) developed allergic diseases, including respiratory (allergic rhinitis and asthma) and skin (atopic dermatitis and eczema) allergies. Lachnospiraceae were detected at significantly higher frequency in allergic infants than nonallergic infants (P < .004); the high fecal count of Lachnospiraceae in allergic subjects appeared at 2 months of age and persisted until 12 months of age. The enrichment of Lachnospiraceae in allergic infants was attributed to the overgrowth of Ruminococcus gnavus, which tended to have a low frequency in nonallergic subjects (P = .0004). Increased R gnavus was observed before the onset of allergic manifestations, and was associated with respiratory allergies (P < .002) or respiratory allergies coexistent with atopic eczema (P < .001). In mice, endogenous R gnavus grew rapidly after sensitization and challenge with ovalbumin. Mice gavaged with purified R gnavus developed airway hyper-responsiveness and had histologic evidence of airway inflammation (asthma). Expansion of R gnavus in mice stimulated secretion of cytokines (interleukin [IL] 25, IL33, and thymic stromal lymphopoietin) by colon tissues, which activated type 2 innate lymphoid cells and dendritic cells to promote differentiation of T-helper 2 cells and production of their cytokines (IL4, IL5, and IL13). This led to infiltration of the colon and lung parenchyma by eosinophils and mast cells.

Conclusions

In a study of a twin cohort (some infants with, some without allergies), we associated development of allergies, particularly respiratory allergies, with increased fecal abundance of R gnavus. Mice fed R gnavus developed airway inflammation, characterized by expansion of T-helper 2 cells in the colon and lung, and infiltration of colon and lung parenchyma by eosinophils and mast cells.