Rotavirus, Intestinal Barrier Function, and Probiotics

See “Lactobacillus rhamnosus GG on Rotavirus-Induced Injury of Ileal Epithelium in Gnotobiotic Pigs” by Liu et al on page 750. To many readers, studies of probiotics in rotavirus infection seem unnecessary given the effectiveness of rotavirus vaccines. Unfortunately, vaccine cost limits widespread immunization and rotavirus infection remains a leading reason for hospitalization and a significant cause of preventable deaths in low-income countries (1). In response to World Health Organization recommendations to include rotavirus vaccines in all immunization programs, public and private collaborations have funded rotavirus vaccinations in developing countries with plans to expand to >30 developing countries by 2015 (2). Even so, many developing countries still will not have rotavirus vaccination programs. Consequently, interest in alternative therapies, such as probiotics, remains keen. The best-studied probiotic, Lactobacillus rhamnosus GG, modestly reduces the duration of diarrhea in acute rotavirus infection and reduces the incidence of health care–associated rotavirus infection through undefined mechanisms (3,4). In this issue of the Journal of Pediatric Gastroenterology and Nutrition, Liu et al (5) evaluate the response of gnotobiotic pigs to rotavirus infection and the effect of Lactobacillus GG on rotavirus infection. During rotavirus infection, cultured intestinal epithelial cells increase mucin production and lose barrier function (6,7). Barrier function depends on intracellular tight junctions to prevent the entrance of microorganisms and other substances into the paracellular space and on adherens junctions to maintain points of cell-to-cell contact (8). Proteins from the zona occludens, occludin, and claudin families contribute to the junctions. Most claudins seal the junctions, whereas claudin-2 increases permeability. Disruption of the intestinal barrier increases expression of claudin-2 and the sealing claudins because the cells repair the tight junctions. In cell cultures, Lactobacillus GG improves barrier function by increasing production of sealing claudins and mucin. To define in vivo events, Liu et al studied gnotobiotic pigs. Pigs develop rotavirus gastroenteritis similar to humans. Additionally, the gnotobiotic status controls for effects of maternal antibodies and commensal microflora. They evaluated 4 treatment groups: mock infection, Lactobacillus GG only, human rotavirus only, and Lactobacillus GG with rotavirus. Lactobacillus GG doses were given before and after infection. They measured clinical symptoms, viral shedding, histology of the ileum, expression levels of tight junction proteins, mucin production, and serum cytokine levels. Unfortunately, the study was underpowered to detect differences in clinical parameters other than a lower incidence of diarrhea in the Lactobacillus GG–fed pigs. Because the probiotics were provided before and after rotavirus infection, the effect could be secondary to prevention or treatment of the infection or both. This design severely limits translation into clinical practice, especially in developing countries where continual delivery of probiotics would be prohibitively expensive. Two other observations provide the major contributions of this report. First, larger inoculums of Lactobacillus GG did not further increase bacterial counts, suggesting the immune system regulates colonization. This observation implies that higher doses may not increase efficacy. Second, Lactobacillus GG treatment altered the expression of tight junction proteins, increased mucin production, and prevented the rise of transforming growth factor-β serum levels. These results provide insight into the mechanisms of rotavirus injury, Lactobacillus GG function, and validate the gnotobiotic pig model. The authors have laid the groundwork for additional studies to define the molecular details of probiotic effects on intestinal epithelium in a model that allows them to test multiple variables such as the commensal microbiome and maternal antibodies. Whether information gleaned from this model ultimately translates to novel treatments for human disease remains speculative.