Mice are in general able to control and contain pulmonary infection with Mycobacterium tuberculosis by the formation of granulomas, whereas a similar mechanism of protection in guinea-pigs and rabbits is less successful, with these structures gradually developing central areas of caseous necrosis and (in the rabbit) even liquefaction of the lesion and cavity formation, similar to that seen in humans with advanced tuberculosis disease. For some time now, the accepted hypothesis to explain these events has been based on an acquired, biphasic T cell response, with protective T cell-mediated immunity activating macrophages to destroy bacilli, and cytolytic T cells mediating delayed-type hypersensitivity leading to the development of caseous necrosis and tissue damage (Dannenberg 1991).
In comparing the mouse and guinea-pig granulomatous response, however, it is clear that lymphocytes accumulate differently in the two species. In the guinea-pig mixtures of lymphocytes and macrophages accumulate in the mantle of the granuloma, and do not associate to any extent with epithelioid macrophages occupying the centre of this structure. Then, when this centre begins to degenerate and become necrotic, the zone of lymphocytes is further distanced from the centre by a layer of foamy macrophages (indicating they have ingested cell debris) and another layer composed of neutrophils (Fig. 1).
In contrast, in the mouse incoming lymphocytes create organized wedges or rafts of cells that are distributed more centrally within the fields of epithelioid macrophages. As these coalesce they may create circlets of lymphocytes, further surrounding any potentially infected macrophages.
It is reasonable to assume that at least a proportion ofthe incoming lymphocytes are CD4+ T cells secreting y-interferon (IFN-y), a key cytokine in tuberculosis immunity (Cooper et al 1993, Flynn et al 1993). In the mouse lesion the juxtaposition of macrophages and lymphocytes suggests that the great majority of epithelioid macrophages in the granuloma will be exposed to reasonable concentrations of this cytokine. (A bacteriostatic rather than bactericidal concentration, given the 'chronic' disease state at this time?)
In the guinea-pig, however, the T cells remain in the mantle. If the granuloma is relatively large, I would propose that it is possible that insufficient cytokine reaches the centre, and as a result the infection eventually reactivates in this region, triggering gradual degeneration and eventual necrosis of the structure.
If this hypothesis is correct, it still does not identify the underlying basis. It could simply involve chemokines; because if cells in the centre are not exposed to IFN-y they fail to produce tumour necrosis factor, which in turn fails to induce local chemokine production (Rhoades et al 1995) and hence fails to attract mononuclear cells into the lesion centre. Other possibilities could reflect the poor invasive properties of guinea-pig T cells in terms of their movement through the extracellular matrix, which may be due to the poor expression of appropriate integrin molecules.
Was this article helpful?