Just how inflamed is the normal gut?
That the science of cartography is limited
- and not simply by the fact that this shading of
forest cannot show the fragrance of balsam,
the gloom of cypresses,
is what I wish to prove
The mammalian intestine has evolved as a tightly regulated and intricate environment concerned with manipulating vast quantities of complex substances. Ingested matter must be processed, transported and appropriately disposed of without stimulating adverse local reactions while contaminating microorganisms with pathogenic potential must be dispatched. Cytokines are of fundamental importance in all of these processes, yet the pleiotropy, interdependence and redundancy of cytokine networks has made this a difficult and often confusing area to chart.
NORMAL INTESTINE: A SITE OF CONTROLLED INFLAMMATION
Cytokine studies in the intestine have, in the past, focused on the possible role of these factors as immunological mediators of inflammatory disease states and it has generally been assumed that activation, as a result of some pathogenic process, was required for their secretion. Thus, inflammatory cytokines, including interleukin (IL) 1, IL-2, tumour necrosis factor (TNF) α and interferon (IFN) γ, have all been claimed to be the primary pathogenic mediators in diseases of the intestine such as Crohn’s disease, ulcerative colitis and coeliac disease.1 However, the paper by Lambrechtset al (see page 643) clearly demonstrates that in the absence of stimulation, the healthy intestine contains significant numbers of IL-4 and IFN-γ elaborating intestinal lymphocytes. Using intracytoplasmic staining followed by flow cytometric analysis, this proportion was found to increase dramatically in response to stimulation with almost two thirds of intestinal lymphocytes expressing IFN-γ. Although this seems at first sight a surprisingly large number of differentiated, IFN-γ secreting T cells, these results do correlate well with other hallmarks of a primed inflammatory immune response which characterises the healthy intestinal tract.2 These include large numbers of CD45RO expressing memory T cells and high expression of the activation markers CD69 and class II molecules. Thus, a majority of resident gut T lymphocytes seem to be terminally differentiated, activated Th1-type T cells primed to secrete IFN-γ. These features raise some interesting questions as to how such large numbers of potential inflammatory cells are normally kept in check. In this respect, the consistent low grade chronicity of inflammatory features found in the disease-free intestine has led to the suggestion that its normal status is “controlled inflammation”.2 Thus the potential to secrete IFN-γ, the key inflammatory cytokine, must be particularly tightly regulated in vivo while remaining readily triggered. Putative key cytokines which play a role in preventing excessive inflammation would include IL-10 and transforming growth factor (TGF) β. Murine studies have suggested that reciprocal IFN-γ and TGF-β expression regulate the occurrence of mucosal inflammation.3 Further studies have elegantly demonstrated this by showing that, when grown in IL-10, ovalbumin specific clones of T cells express IL-10, IFN-γ and TGF-β but not IL-2 or IL-4. These clones could then prevent colitis when transferred into a murine model—interestingly, only after feeding with OVA protein.3 However, caution needs to be exercised when extrapolating from mouse to man—for example, in the mouse only a small frequency of normal murine intestinal lymphocytes are differentiated to produce IFN-γ.4 This contrasts with the human studies of Lambrechts et al where almost two thirds of intestinal lymphocytes expressed IFN-γ when stimulated. Such strong differences suggest that different networks might operate in each species.
ADVANTAGES OF THE “CONTROLLED INFLAMMATORY” STATUS
The fundamental features of the gastrointestinal tract impose a need for a range of protective mechanisms primed to deal with the potentially harmful challenges inevitably encountered during food consumption. Indeed, it is likely that this was a primary driving force for the evolution of an immune system. For humans, such mechanisms would have had distinct survival advantages when food was gathered in the wild and eaten uncooked or in unhygenic conditions. The lack of appropriate immunological stimulation as a result of eating almost sterile food in the obsessively hygienic conditions characteristic of the Western way of life may contribute to the current increase in food intolerance and associated allergies. It has already been proposed that the absence of infection may contribute to the increasing incidence of asthma and atopy.5 This recent study described an inverse association between tuberculin responses and atopic disorders in 1000 Japanese school children. It was proposed that infection, which stimulated a Th1-type environment with IL-12, IFN-γ and TNF-α secretion, might protect against Th2 mediated asthma and atopy by inhibiting Th2 type cytokines. Conceivably, under modern living conditions and the absence of gastrointestinal infections, the system defaults to a predominant Th2 mode in certain individuals, thus promoting the chances of Th2 mediated diseases such as food allergy.
CYTOKINE ELABORATION AND INFLAMMATORY BOWEL DISEASE
If the normal intestine is constantly in a state of controlled inflammation, it is easy to speculate how breakdown of control may contribute to the aetiopathogenesis of inflammatory bowel disease. Changes in genetic regulation of cytokine expression may significantly influence the composition of the local cytokine milieu and thus contribute to the propensity to develop inflammatory disease. There is increasing evidence that polymorphisms of the promoter regions of cytokine genes can lead to altered expression patterns and thus influence immune responses.6 Thus, a genetically influenced local environment which was rich in IFN-γ or TNF-α could lead to triggering or overpriming of the system. Alternatively, low levels of IL-10 or TGF-β could lead to breakdown in immunoregulation. Indeed, artificial interference with regulatory cytokine pathways using gene knockout technology has, in many cases, resulted in animal forms of chronic intestinal inflammation which resemble human inflammatory bowel disease.7 Moreover, recent studies have shown that a single infusion of a chimeric monoclonal anti-TNF antibody was an effective short term treatment in many patients with moderate to severe Crohn’s disease.8 However, despite such encouraging results, the complexity of cytokine networks and their overlapping roles leads one to intuit that the notion that interference with a single cytokine could cure all cases of inflammatory bowel disease may be naive and possibly dangerous in the long term.
Localisation of cytokine production to individual cell populations and different stages of stimulation will be fundamental to understanding how they might be ultimately manipulated in pathogenic situations. It is becoming increasingly clear that intestinal T cells are highly heterogeneous with regard to TCR and co-receptor expression, activation status and origin.9-11 It is likely that their cytokine profiles will be equally individual. Lambrechts et al found that virtually all IL-4 and IFN-γ secreting cells in the lamina propria were CD4+. In the epithelium, IFN-γ elaboration was found in CD8+ populations as well as CD4+. Further isolation of various subpopulations of lymphocytes from both intestinal compartments and comprehensive analysis of their cytokine production potential will be required in order to understand their function. TNF, IL-12, IL-18, and TGF-β will be particularly interesting cytokines to study from this point of view. Cytokine production by cells other than lymphocytes will also have important functional implications—for example, it already has been shown that epithelial cells elaborate IL-6, IL-10 and, more recently, IL-7.12 Development of the technical ability to analyse coexpression of two or more cytokines and changes in relative proportions of the cytokine cocktail produced by individual subpopulations will lead to more accurate mapping of cytokine networks and their influences. However, a two-dimensional fixed chart could only ever be a limited representation of this plastic, interdependent system. Some sort of a multidimensional, dynamic map will be required to make accurate predictions of how the intestinal immune response might be manipulated by influencing individual cytokine expression. Thus, computer modelling of cytokine networks may lead to successful prediction of the small changes in individual components of cytokine networks required for disease treatment and prevention while maintaining homoeostasis of the gastrointestinal immune system.