Article Text

Intolerance of the dirty intestine
  2. K J MALOY
  1. Institute of Experimental Immunology
  2. Universitätsspital
  3. Schmelzbergstrasse 12
  4. CH 8091 Zürich
  5. Switzerland
  6. Department of Medicine
  7. GKT School of Medicine
  8. Bessemer Road
  9. London SE5 9PJ, UK

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

See article on page 819

The mucosal immune system faces the delicate task of co-existence with a luxuriant commensal intestinal bacterial flora (1012 bacteria/g faeces in the colon, and roughly 103 different species, with anaerobes predominating). Yet a protective immune response to invasive enteric pathogens is also mandatory. Of course, any commensal organism can become a pathogen in appropriate circumstances, and the magnitude of this balancing act is illustrated by the similarity between proteins of the harmless commensal Escherichia coli and its pathogenic derivatives (or the Shigellagenus). The essential differences between innocent and harmful bacteria reside in toxin production and qualities of adherence to, or penetration of, the intestinal epithelial cell layer.

The barrier function of the intestine is clearly disrupted in active inflammatory bowel disease (IBD), with evidence for greater systemic penetration by commensal organisms. Whether abnormal intestinal permeability precedes intestinal inflammation in individuals predisposed to IBD, is still unclear, although it is a major feature of Crohn’s disease relapse. The “passive” intestinal barrier (of the epithelial cell layer linked by junctional complexes and overlaid with mucus) is certainly not (in real life) as fixed as one might suppose.1 We constantly face insults to barrier function, either through our own actions (ingestion of non-steroidal anti-inflammatory drugs, alcohol) or from intestinal pathogens. Epithelial cells are also actively involved in defence through chemokine and cytokine release. It is easy to see how a vicious cycle following penetration by luminal bacteria or their products might arise, ending in chronic IBD, yet most of us avoid it.

Over the past six years the study of genetically manipulated rodents has contributed enormously to the understanding of the circumstances which predispose to intestinal inflammation. Ablating the function of a large number of different immunological genes (including interleukin (IL) 2, IL-10 or αT cell receptor) or inserting HLA-B27 each independently renders the animal liable to develop spontaneous intestinal inflammation that may usually be attenuated or avoided by breeding and keeping the animals in very clean (SPF) or germ-free facilities. Although genetic loci linked to human IBD have been described, the hunt for the genes themselves continues, so many of the animal genetic abnormalities may be somewhat artificial. Nevertheless, they do provide support, in well defined conditions, for the concept that upsetting the delicate balance between the mucosal immune system, the epithelial cell layer and the commensal bacterial flora results in chronic intestinal inflammation.

Duchmann and coworkers, in this issue (see page 812), have examined the reactivity of T cell clones, derived from IBD intestinal mucosa, against commensal bacteria. It is clear that the mucosa of active Crohn’s disease contains an increased proportion of activated T cells,3 and T cell cloning has generally proved a powerful immunological technique as it provides a culture of T cells with a single receptor with specificity for short (9–15 amino acid) peptide epitopes. From such clones the major antigenic determinants for helper and cytotoxic T cells in viral and bacterial infections have been elucidated.

In general, working out the specific antigens to which CD4+ T cells in the intestinal mucosa respond is a formidable task, complicated by their very poor antigen specific proliferative responses in culture and the diversity of the luminal antigen mixture. Nevertheless, Duchmann and colleagues have previously presented data that T cells isolated from the intestinal mucosa of control subjects will proliferate in vitro in response to relatively crude fractions of bacteria isolated from the intestinal (heterologous) flora of a different individual, but not from their own flora. Patients with Crohn’s disease, however, have intestinal T cells capable of responding to their own (autologous) flora. Thus Crohn’s disease could be interpreted as a failure of mucosal tolerance to the indigenous flora, an idea in keeping with the data from the animal models and with the clinical effectiveness of faecal stream diversion. Despite the diversity of the human intestinal microflora, there is considerable homology between proteins of related species and common carriage of many species by different individuals, so the differences that bring about responses to the heterologous flora in normal subjects are still unclear.

An important technical issue surrounds the way in which the T cell clones were derived in this paper. It is difficult to get T cells from the intestinal mucosa to proliferate well in response to antigen, so to produce clones of identical cells the stimulation process had to be non-specific—phytohaemagglutinin (PHA) followed by expansion on irradiated allogenic feeder cells. Although the idea is to obtain representative clones, the responses to bacterial sonicates may not reflect the antigen specificities of the initial T cells. With this caveat, there are three main results. Firstly, there was considerable cross reactivity in the response of CD4+ clones to anaerobic (Bifidobacterium andBacteroides) and aerobic enterobacteria. Secondly, the authors show that their previous observation whereby T cells from patients with IBD respond to crude preparations of the autologous flora, can be applied (in some cases) at the level of T cell clones. Thirdly, they analysed which bacterial species within a heterologous mixed isolate could stimulate a T cell clone from a patient with ulcerative colitis and showed that aerobic enterobacteria were mainly responsible and curiously some colonies of a bacterial species (e.g. E coli) might stimulate this clone whereas others would not.

So, on the one hand, there seems to be cross reactivity in the proliferative responses of T cell clones from patients with IBD between different bacterial species, and, on the other, the responses to the heterologous flora involve many common aerobic species. Where is the window in which “tolerance” to the autologous flora develops or collapses at the T cell level? A substantial amount of work will be needed to unravel this question. The beauty of T cell clones is that specificities to individual protein molecules (or other structural bacterial components) can be determined (if bacterial proteins are first purified), and this could sort out the molecular basis of cross reactivity and detect the window(s) in the autologous/heterologous flora in individual cases. Unfortunately each person is likely to be different because of the diversity of MHC class II in the human population which presents antigenic peptide to CD4+ T cells.

Duchmann et al’s paper is an important contribution to the evidence that difficulties in handling the commensal microflora are the key to IBD. Yet the mucosal immune system is sophisticated, and in addition to the presence or absence of lamina propria CD4+ T cells that respond to commensal bacterial determinants (as investigated here), there are many other levels of regulation, including unresponsive T cells and those which produce downregulatory cytokines but do not proliferate.4 ,5 The relative contributions of these mechanisms in health and their defects in IBD are still not known.


Linked Articles