Cellular Microbiology (2001) 3(1), 1±11 Microreview The gastrointestinal ecosystem: a precarious alliance among epithelium, immunity and microbiota Vance J. McCracken 1 and Robin G. Lorenz 1,2 1 Department of Pathology and Immunology, Washington University School of Medicine, St. Louis, Missouri 63110, USA. 2 Department of Medicine, Washington University School of Medicine, St. Louis, Missouri 63110, USA. Summary The gastrointestinal (GI) tract is a complex ecosystem generated by the alliance of GI epithelium, immune cells and resident microbiota. The three components of the GI ecosystem have co-evolved such that each relies on the presence of the other two components to achieve its normal function and activity. Experi- mental systems such as cell culture, germ-free animal models and intestinal isografts have demonstrated that each member of the GI ecosystem can follow a predetermined developmental pathway, even if isolated from the other components of the ecosystem. How- ever, the presence of all three components is required for full physiological function. Genetic or functional alterations of any one component of this ecosystem can result in a broken alliance and subsequent GI pathology. A more detailed understanding of the inter- actions among microbiota, GI epithelium and the immune system should provide insight into multiple human disease states. Introduction The adult gastrointestinal (GI) ecosystem comprises a stable alliance among the resident microbiota, immune mediators and the epithelial barrier. Although each con- stituent of this alliance is capable of a certain degree of development independent of the other members, all three components are essential for complete functional and developmental maturity of the GI ecosystem. This mini- review will focus on the involvement of the resident GI microbiota in the development of intestinal epithelial morphology and immune system functions. We will use the term `resident' microbiota to refer to microbial species found in the healthy gut, without reference to evolutionary implications, commensal relationships or potential for opportunistic pathogenesis. Interactions between bacteria and animals have pri- marily been studied in terms of pathogenic relationships. However, it is apparent that bacteria exert beneficial as well as deleterious effects on their animal hosts. The mammalian GI tract harbours a dense and diverse com- munity of bacteria, containing up to 1 Â 10 14 colony- forming units (CFUs) of bacteria, some 10 times the amount of host cells (Berg, 1996). The host is protected from the resident intestinal microbiota by the physical and chemical barriers formed by the GI epithelium (Kagnoff and Eckmann, 1997; Bevins et al., 1999; O'Neil et al., 1999). These barriers are reinforced by the acquired mucosal immune response (McGhee et al., 1999). During the development of the ecosystem, antigen-specific B and T cells learn to regulate their responses to resident micro- biota, such that a state of controlled inflammation is established in adult hosts (Cebra, 1999). If any com- ponent of the ecosystem is missing, or genetically or environmentally altered, the precarious alliance is broken and pathology ensues. Members of the alliance Resident microbiota Analysis of intestinal microbial ecosystems is complicated by the complex nature of intestinal microbial communities. The number of bacterial species has been estimated to be near 400, although newer studies indicate that a smaller number of species comprise the majority of the intestinal microbial species (Moore and Holdeman, 1974; Vaughan et al., 2000). Until recently, microbial ecologists relied largely on techniques that required cultivation of organisms on selective media. The use of selective media specific for different types of bacteria has the disadvantage of being time- and labour-intensive and also imposes an a priori bias on the types of bacteria that can be enumerated. Additionally, only 20±40% of bacterial species from mammalian gastrointestinal tracts can be cultured and Q 2001 Blackwell Science Ltd Received 28 June, 2000; revised 4 August, 2000; accepted 10 August, 2000. *For correspondence. E-mail lorenz@pathbox.wustl. edu; Tel. (11) 314 362 3669; Fax (11) 314 747 2048.