The finding that intestinal viruses can substitute for intestinal bacteria to promote the health of their mammalian hosts raises the possibility that viruses in the gut may be beneficial in some circumstances. See Letter p.94
A dense population of microorganisms inhabits the intestinal tract, including bacteria, archaea, fungi and viruses. Intestinal bacteria benefit the host by aiding nutrition, promoting immune-cell development and protecting from intestinal damage, but it is unclear whether other members of the microbiota have similar roles. Viruses are relatively poorly studied members of the intestinal microbial community, and mammalian intestinal viruses are generally thought to be harmful or neutral to the host. But on page 94 of this issue, Kernbauer et al.1 show that, in the absence of bacteria, mammalian intestinal viruses promote gut homeostasis and protect the intestine from injury and from pathogenic bacteria. Their work raises the possibility that mammalian intestinal viruses may benefit the host in some instances.
The collection of viruses in the intestine — the intestinal virome — includes viruses that infect bacteria (bacteriophages), archaeal viruses, plant viruses and mammalian viruses2,3,4. Although bacteriophages are abundant and ever-present inhabitants of the intestine, mammalian viruses are detectable only some of the time. Members of the mammalian virome include viral pathogens that persist after a disease has ended, non-pathogenic viruses that are often detected in healthy individuals, and uncharacterized viruses whose similarity to known viruses in terms of nucleic-acid sequence is low5,6. Murine norovirus (MNV) is a relatively common member of the mouse intestinal virome and several MNV strains have been discovered in animal-research facilities. MNV is generally non-pathogenic in mice that have a functioning immune system, but can cause disease in some immune-deficient mice. Kernbauer and colleagues used three MNV strains as representative members of the mouse intestinal virome to determine whether viruses can influence intestinal homeostasis in 'germ-free' mice that lack a microbiota.
Germ-free mice are microbiologically sterile, and because of the absence of beneficial bacteria in their intestines they have abnormal intestinal morphology and defective development of immune cells called lymphocytes7. Antibiotic-treated mice also have several of these abnormalities. The authors found that MNV infection could repair many of the morphological and immunological defects in these mice (Fig. 1). For example, it restored the thickness of the villi (projections of the mucous membrane that lines the intestine) and the number of granules in the Paneth cells of the intestinal lining, which are filled with antimicrobial compounds. Also restored were the numbers of CD4+ and CD8+ T cells (two classes of lymphocyte), the production of immune-stimulatory molecules by these cells, more-normal levels of antibodies, and a balanced population of innate lymphoid cells, another class of intestinal immune cell. Thus, the presence of a single type of virus reversed many of the defects caused by the lack of bacteria in germ-free or antibiotic-treated mice.
MNV-induced restoration of intestinal cell numbers, cell function and tissue architecture in germ-free mice suggests that the virus can replace some functions of the microbiota. But are there major functional consequences of these MNV-associated changes? For example, can viral infection of germ-free mice limit disease and aid host survival in the face of injury or infection with pathogenic bacteria? To test this, Kernbauer and colleagues treated normal and antibiotic-treated mice with an intestine-damaging chemical and monitored their survival in the presence or absence of MNV infection. The intestinal injury killed more antibiotic-treated mice than conventional mice, but MNV infection enhanced the survival of antibiotic-treated mice. Similarly, infection with pathogenic bacteria was more damaging in antibiotic-treated mice than in conventional mice, but viral infection reduced disease in antibiotic-treated mice. These data demonstrate that MNV infection can functionally substitute for the microbiota to enhance the animals' health and survival.
This work raises many interesting questions. Mammalian intestinal viruses are generally thought to be detrimental to the host, but previous studies have shown beneficial effects of viral infection in other circumstances. For example, herpesviruses can suppress bacterial infections, and retroviruses were involved in the evolution of the placenta8,9. This work provides the first evidence for beneficial effects of a mammalian intestinal virus, but other intestinal viruses may exist in commensal associations, in which the virus can replicate without harming the host, or in mutualist associations, in which both host and virus benefit. Further investigations will be needed to determine whether such viruses are frequent inhabitants or transient members of the microbiota, and which mammals harbour these viruses and to what effect.
Another outstanding question is, do mammalian intestinal viruses benefit the host in the context of the 'normal' microbiota? The current study examined the beneficial effects of MNV using germ-free mice or mice treated with multiple antibiotics. Presumably, the beneficial effects of a complete microbiota would mask the beneficial effects of a viral infection. However, the presence of a single type of bacterium, the segmented filamentous bacteria, in the microbiota is known to influence pro-inflammatory responses10, and particular intestinal viruses might induce similarly specific responses even in a healthy host with a normal microbiota. Indeed, Kernbauer et al. found that the three closely related MNV strains induced slightly different responses in the germ-free mice.
Finally, could mammalian intestinal viruses be useful as probiotics? Intestinal inflammation in humans can be caused by antibiotic use, genetic background and other factors. In some cases, such disease can be reversed by faecal transplant — and the consequent transfer of the intestinal microbiota — from a healthy donor. Perhaps certain mammalian intestinal viruses could also ameliorate disease. Moreover, it is possible that some of the observed benefits following faecal transplants are due to viruses present in the donor samples. Future work exploring the intestinal virome should provide some answers to these questions.