Commentary: Remembrance of microbes past.

Sigmund Freud (1856–1939) established a central theme of psychoanalysis: the past is alive in the present. Influences experienced in the past are not something that can be completely outgrown by the individual or society; they remain vital parts of existence. Rene Dubos (1901–1982), French-born American microbiologist, experimental pathologist, environmentalist, humanist, and Pulitzer Prize-winning author, explored in a number of books the interplay between environmental forces and the physical, mental, and spiritual development of humankind. His article in the journal Pediatrics entitled ‘Biological Freudianism: lasting effects of early environmental influences’ and written in collaboration with his postdoctoral fellows Dwayne Savage and Russell Schaedler (soon to be eminent scientists in their own right) encapsulated this theme.1 Drawing on results obtained from experiments with specificpathogen-free mice, the authors concluded that ‘From all points of view, the child is truly the father of the man, and for this reason we need to develop an experimental science that might be called biological Freudianism. Socially and individually the response of human beings to the conditions of the present is always conditioned by the biological remembrance of things past’. With this two-sentence statement the works of William Wordsworth, Sigmund Freud, and Marcel Proust were deftly wedded to biology. Only touched upon in the Pediatrics article was the pioneering work of the Dubos group concerning the impact of the diverse collection of bacterial species (the microbiota) that inhabit the gut of mice from soon after birth. As recalled by Dwayne Savage, the NCS (New Colony Swiss) mouse colony derived at The Rockefeller University was the first murine colony in the world in which the animals, while harbouring a gut microbiota, were nevertheless free of certain mouse pathogens and could be bred in sufficient quantity for major experiments.2 Dubos and colleagues soon realized that NCS mice differed in several characteristics from SS mice (Standard Swiss, from which the NCS colony had been derived). NCS mothers bore on average more infants per litter, and their offspring grew faster and were larger than SS mice (even when fed diets low in lysine and threonine). It was suggested that the differences between NCS and SS mice were the result of mutations (in other words, that a new mouse strain had been selected); however, astoundingly, NCS mice housed with SS mice reverted to the characteristics of SS mice in all properties tested. The faecal microbiota of the two colonies of mice were different when analysed by bacteriological culture. Unlike SS mice, NCS animals did not harbour facultatively anaerobic gram-negative bacteria such as Escherichia coli among the members of their gut microbiota. Remarkably, NCS mice were tolerant to parenteral doses of endotoxin that were lethal within a few days of administration to SS mice. Exposure of NCS animals to SS mice during early life, or prior injections with small doses of heat-killed cells of gram-negative bacteria that thus exposed the animals to sublethal amounts of endotoxin, increased the susceptibility of NCS mice to match that of SS animals. Dubos and colleagues concluded that the enterobacteria, present in relatively high numbers in the gut of infant SS mice, sensitized the animals so that, as adults, they were highly susceptible to endotoxin. These observations led Dubos and colleagues to devote almost 20 years to the study of gut microbiota composition and gut microbial ecology. They established experimental methods and observations that, as Savage has appropriately stated, ‘should endure in history’.2 Many clues to the influences of bacteria on the mammalian host have been obtained from comparisons of the biochemical and physiological characteristics of germ-free (raised in the absence of demonstrable microbes) and conventional (colonized by microbes) animals.3 The impact of even a single bacterial species on the host animal can also be ascertained using this approach. Gnotobiotic (‘defined biota’) work like this can now be done at a sophisticated level because of the availability of advanced imaging technology, as well as genome sequences of experimental animal species and the consequent preparation of DNA microarrays that can be used to measure gene expression. The potential for obtaining exciting knowledge of the mechanistic influences of gut microbiota on a host using this approach has been demonstrated by the work of Gordon and colleagues, who have studied the impact of colonization of formerly germ-free mice by the bacterial species Bacteroides thetaiotaomicron.4 Perhaps most striking have been their observations concerning angiogenesis in the murine gut. Quantitative three-dimensional imaging studies showed that a plexus of branched and interconnected blood vessels developed postnatally in small bowel villi of conventional mice. Angiogenesis coincided with the establishment of the gut microbiota. Vascular development was arrested in germ-free mice, but could be restarted by colonization of the gut by a conventional gut microbiota or by B. thetaiotaomicron, which had been shown in other experiments to up-regulate expression of the murine angiogenin-3 gene in the ileal mucosa. Other observations showed that interaction between the gut microbiota and Paneth cells was essential for the regulation of angiogenesis.4,5 The gut of newborn human infants resembles those of germfree animals because it is not yet colonized by a microbiota. This germ-free state is short-lived, however, because within minutes of birth bacteria in the faeces that have been involuntarily expelled by the mother during labour, together with environmental microbes, have the opportunity to colonize the Department of Microbiology and Immunology, University of Otago, PO Box 56, Dunedin, New Zealand. E-mail: [email protected]