Gut-level decisions in peace and war

The gut contains most of the immune cells in the body, yet fights off bacteria only if they are pathogenic, leaving commensal microbes largely unscathed. Two studies examine the basis for this effect.

The intestine has developed the largest border surface of the body to digest and absorb nutrients. The gut also has complex interactions with beneficial colonizing microorganisms, which reside mostly in the distal portion of the intestine. These microbes facilitate the metabolism of nutrients and the development of the mucosal immune system1.

The gastrointestinal tract is also exposed to environmental threats, mainly pathogens that have developed strategies to penetrate the intestinal barrier. Therefore, the mucosal immune system has the difficult task of limiting inflammatory responses to commensal bacteria while retaining the ability to initiate protective adaptive immune responses to pathogens2. How can this be achieved?

Two recent studies in the Journal of Clinical Investigation3 and in Science4 suggest that the intimate interplay between the mucosal microenvironment and immune cells regulates the induction of inflammatory responses and immunity to intestinal bacteria.

It is becoming clear that epithelial cells modulate the mucosal immune function. A key antigen-presenting cell, the dendritic cell, normally can extend dendrites, like periscopes, into the intestinal lumen to take up bacteria5. Niess et al.4 show that this uptake is regulated by CX3CL1, a chemokine produced by intestinal epithelial cells.

Smythies et al.3 describe a population of macrophages with 'inflammatory unresponsiveness' or 'anergy,' which is conferred by factors derived from the cells of the stroma—fibroblasts and other nonspecialized cells. These macrophages are phagocytic, but are unable to initiate inflammatory responses to bacteria.

Two major mechanisms of bacterial uptake in the gut have been described. In the first, microfold cells in the gut serve to transport fluids, nutrients and bacteria5, including pathogens that have learned to exploit this cell type. The second mechanism is mediated by dendritic cells scattered throughout the intestinal epithelium that snatch bacteria from the lumen by extending their dendrites into it5. The integrity of the epithelial barrier is preserved during this process because dendritic cells can establish new tight junction–like structures with adjacent epithelial cells5.

Salmonella: fodder for dendritic cells Credit: CAMR, AB Dowsett, Science Photo Library

To show that the chemokine CX3CL1 regulates this sampling, Niess et al. used mice in which the gene encoding CX3CR1, the receptor for CX3CL1 was replaced on one or both alleles with the cDNA encoding green fluorescent protein. They examined fluorescent dendritic cells in mice heterozygous for or deficient in CX3CR1, and found that the extension of dendritic cell protrusions into the intestinal lumen required CX3CR1. In the receptor-knockout mice, the dendritic cells were recruited into the gut, but were unable to extend dendrites across epithelial cells.

The researchers went on to show that intercalating CX3CR1 dendritic cells are involved in the uptake of S. typhimurium that are both invasive and noninvasive (unable to penetrate epithelial cells). Further, in the absence of CX3CR1, noninvasive bacteria were unable to cross the intestinal barrier. This observation argues against a contribution by other subsets of dendritic cells and villous microfold cells7 in the uptake of noninvasive bacteria outside of the Peyer patches, specialized inductive sites of the mucosal immune response.

The authors next showed that intercalating CX3CR1-expressing dendritic cells are located mainly in the terminal region of the small intestine, where CX3CL1 is expressed and where the gradient of intestinal bacteria gradually increases; this observation suggests that the bacteria may facilitate dendrite extension.

The authors found that the CX3CR1 knockout mice had enhanced susceptibility to infection with S. typhimurium, but it is difficult to conclude that this occurs because of the inability of CX3CR1-deficient dendritic cells to sample lumenal bacteria. In fact, CX3CR1-deficient dendritic cells, and presumably macrophages that express CX3CR1 as well, also have defects in phagocytosis of S. typhimurium.

Although CX3CR1 is required for the extension of dendritic cell protrusions, exactly how the molecule operates is unknown. One possibility is that CX3CR1 engagement with its ligand could modulate the expression of tight-junction proteins that are required for dendritic cells to creep between epithelial cells8. Finally, it would be interesting to analyze the origin of intestinal dendritic cells expressing CX3CR1. It is unlikely that they are derived from monocytes resident in the gut, as these cells require CX3CR1 for homing to noninflamed tissues9 (although it is possible that the gut has different homing properties.)

Because both the microfold cell– and dendritic cell–mediated mechanisms target noninvasive bacteria as well as pathogens, how is the generation of inflammatory responses regulated in the gut? Smythies et al. begin to fill this gap. They describe a unique phenotype of intestinal resident monocyte-derived macrophages. They show that these macrophages do not express receptors for the innate immune response, including the coreceptor for lipopolysaccharide (CD14), Fc receptors (FcαR, and FcγR I, II and III) and complement receptors (CR3 and CR4). But these macrophages are perfectly competent at phagocytosis and bacterial killing. It also seems that the macrophages have a broad inability to release cytokines, including proinflammatory mediators such as tumor necrosis factor-α, interleukin (IL)-6 and IL-10.

It is clear that intestinal macrophages differ substantially from the blood monocytes from which they derive. This suggests that the mucosal microenvironment can influence the differentiation properties of monocytes into tissue macrophages. Indeed the authors show that stromal-derived factors and in particular TGF-β can confer a similar 'noninflammatory' phenotype to blood monocytes, inducing the downregulation of innate receptors and cytokine release.

The report by Smythies et al. nicely complements recent studies showing that dendritic cells in the mucosa display unique functions. Such dendritic cells preferentially induce mucosal immune responses and confer gut tropism to T cells10. Moreover, similar to intestinal macrophages, gut-resident dendritic cells are conditioned by factors derived from the epithelial cells to become 'noninflammatory' dendritic cells (unable to release IL-12 and to activate inflammatory T helper type 1 (TH1) cells in response to TH1-promoting pathogens (M.Rimoldi et al., unpublished data).

All together, these data strongly suggest that in the 'steady' or unperturbed state mucosal inflammatory cells acquire a profound and irreversible condition of 'inflammatory anergy' that limits mucosal inflammation and maintains gut immune homeostasis (Fig. 1). Deregulation affecting the establishment of inflammatory unresponsiveness may promote the inflammation associated with inflammatory bowel disease like Crohn disease. Under normal conditions it is unlikely that resident CX3CR1 dendritic cells or macrophages can initiate inflammatory responses to commensal bacteria or to pathogens, during the early phases of infection.

Figure 1: Cell dynamics under steady-state conditions and during Salmonella typhimurium infection.

Ann Thomson

TGF-β and other factors derived from stromal cells such as fibroblasts prompt resident intestinal macrophages to acquire a profound inflammatory anergy or 'unresponsiveness.' These noninflammatory macrophages do not express innate response receptors and do not release inflammatory cytokines in response to bacteria—but they retain phagocytic and bacteriocidal activity. Similarly, factors derived from epithelial cells condition intestinal dendritic cells to become noninflammatory—although these CX3CR1-expressing cells are still able to open the tight junctions and extend dendrites into the intestinal lumen for uptake of commensal bacteria. This helps maintain the homeostasis of the gut. During infection, invasive bacteria activate epithelial cells to produce proinflammatory mediators (such as CXCL-8, CCL-20 and CCL-2) that recruit additional immune cells including neutrophils; this leads to an increased number of intercalating CX3CR1-expressing dendritic cells. Blood vessels just underneath the epithelial layer are the perfect site to allow fast recruitment of inflammatory cells. The dendritic cells and monocytes newly recruited to the infected site probably encounter an inflammatory state which promotes their differentiation into inflammatory cells.

The situation is different later in infection when invasive bacteria like Salmonella are sensed by epithelial cells that release proinflammatory mediators and chemokines (such as MCP-1 (CCL-2), IL-8 (CXCL-8) and MIP-3α (CCL-20)2,12) which attract monocytes, neutrophils and dendritic cells into the infected area. This movement is facilitated by the close proximity between blood vessels and epithelial cells, allowing fast recruitment of inflammatory cells at the site of bacterial entrance.

Nonconditioned, newly recruited immune cells from the blood may find an immunostimulatory environment in the infected site, generated by epithelial-derived factors that triggers differentiation of these cells into inflammatory cells. Therefore, the initiation of protective responses to bacteria depends on the ability of recruited immune cells to differentiate into inflammatory cells before they are conditioned and 'tamed' by tissue-derived factors.


  1. 1

    Macpherson, A. J. & Harris, N. L. Nat. Rev. Immunol. 4, 478–485 (2004).

    CAS  Article  Google Scholar 

  2. 2

    Mowat, A. M. Nat. Rev. Immunol. 3, 331–341 (2003).

    CAS  Article  Google Scholar 

  3. 3

    Smythies, L. E. et al. J. Clin. Invest. 115, 66–75 (2005).

    CAS  Article  PubMed  Google Scholar 

  4. 4

    Niess, J. H. et al. Science 307, 254–258 (2005).

    CAS  Article  Google Scholar 

  5. 5

    Rescigno, M. et al. Nat. Immunol. 2, 361–367 (2001).

    CAS  Article  Google Scholar 

  6. 6

    Neutra, M. R., Mantis, N. J. & Kraehenbuhl, J. P. Nat. Immunol. 2, 1004–1009 (2001).

    CAS  Article  Google Scholar 

  7. 7

    Jang, M. H. et al. Proc. Natl. Acad. Sci. USA 101, 6110–6115 (2004).

    CAS  Article  Google Scholar 

  8. 8

    Rimoldi, M., Chieppa, M., Vulcano, M., Allavena, P. & Rescigno, M. Ann. NY Acad. Sci. 1029, 1–9 (2004).

    Article  Google Scholar 

  9. 9

    Geissmann, F., Jung, S. & Littman, D. R. Immunity 19, 71–82 (2003).

    CAS  Article  Google Scholar 

  10. 10

    Kelsall, B. L. & Rescigno, M. Nat. Immunol. 5, 1091–1095 (2004).

    CAS  Article  Google Scholar 

  11. 11

    Sansonetti, P. J. Nat. Rev. Immunol. 4, 953–964 (2004).

    CAS  Article  PubMed  Google Scholar 

Download references

Author information



Rights and permissions

Reprints and Permissions

About this article

Cite this article

Rescigno, M., Chieppa, M. Gut-level decisions in peace and war. Nat Med 11, 254–255 (2005).

Download citation

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing