The immunopathology of tuberculosis

Immunity to M. tuberculosis requires a T helper 1 (Thl) pattern of response accompanied by several types of cytotoxic T cell (Flynn et al 1993,1995a,b, Rook & Hernandez-Pando 1996). A Th2 response is disastrous: even a small Th2 component exacerbates the infection in mice (Lindblad et al 1997, Rook & Hernandez-Pando 1996). Thus, if the animals are pre-immunized so that they have a 'pure' Th1 response to culture filtrate antigens of M. tuberculosis, or to the common antigens present in a distantly related environmental saprophyte, they are protected (Figs 1 & 2). However, if they are preimmunized so that they have a mixed Th1+Th2 pattern of response to these common antigens, they have increased susceptibility to pulmonary tuberculosis manifested as accelerated death (Fig. 3a), increased percentage of the lung affected by pneumonia (Fig. 3d) and accelerated appearance of Th2 cells in the lesions (Fig. 3b) (Hernandez-Pando et al 1997, Rook & Hernandez-Pando 1996).

FIG. 1. Unimmunized mice were infected with Mycobacterium tuberculosis H37Rv by the intratracheal route, and the following parameters studied at intervals for 120 days. (a) Percentage survival. (b) The percentage of interleukin (IL)-2-positive (□) or IL-4-positive (#) lymphocytes in the zones of perivascular inflammation, as detected by immunohistochemistry. (c) The delayed hypersensitivity response (foot-pad test) to soluble antigen of M. tuberculosis, 24 h after challenge (□). After reading the swelling, 1 ^g of tumour necrosis factor a (TNF-a) was injected into the same site, and swelling was read again 20 h later (•). (d) Percentage of the lung affected by pneumonia. The pneumonic zones always contained numerous IL-4-positive cells. Data adapted from Hernandez-Pando et al (1997). Error bars represent S.D.

FIG. 1. Unimmunized mice were infected with Mycobacterium tuberculosis H37Rv by the intratracheal route, and the following parameters studied at intervals for 120 days. (a) Percentage survival. (b) The percentage of interleukin (IL)-2-positive (□) or IL-4-positive (#) lymphocytes in the zones of perivascular inflammation, as detected by immunohistochemistry. (c) The delayed hypersensitivity response (foot-pad test) to soluble antigen of M. tuberculosis, 24 h after challenge (□). After reading the swelling, 1 ^g of tumour necrosis factor a (TNF-a) was injected into the same site, and swelling was read again 20 h later (•). (d) Percentage of the lung affected by pneumonia. The pneumonic zones always contained numerous IL-4-positive cells. Data adapted from Hernandez-Pando et al (1997). Error bars represent S.D.

FIG. 2. Mice were immunized once with the optimally T helper 1-inducing dose (107 heat-killed bacilli) of the environmental saprophyte Mycobacterium vaccae. Two months later they were infected with Mycobacterium tuberculosis H37Rv by the intratracheal route. The parameters studied are identical to those described in the legend to Fig. 1, with which these data should be compared. Data adapted from Hernandez-Pando et al (1997). Error bars represent S.D.

FIG. 3. Mice were immunized once with a 100-fold excessive dose (109 heat-killed bacilli) of the environmental saprophyte Mycobacterium vaccae. This evokes a mixed T helper 1 (Th1)+Th2 pattern of response. Two months later they were infected with Mycobacterium tuberculosis H37Rv by the intratracheal route. The parameters studied are identical to those described in the legend to Fig. 1, with which these data should be compared. Data adapted from Hernandez-Pando et al (1997). Error bars represent S.D.

FIG. 3. Mice were immunized once with a 100-fold excessive dose (109 heat-killed bacilli) of the environmental saprophyte Mycobacterium vaccae. This evokes a mixed T helper 1 (Th1)+Th2 pattern of response. Two months later they were infected with Mycobacterium tuberculosis H37Rv by the intratracheal route. The parameters studied are identical to those described in the legend to Fig. 1, with which these data should be compared. Data adapted from Hernandez-Pando et al (1997). Error bars represent S.D.

The Th2 component in human tuberculosis

There is strong evidence that progressive human disease also has an inappropriate Th2 component, though its manifestation is more subtle than in the murine model. The peripheral blood of tuberculosis patients contains Th2 cells that respond to mycobacteria by expressing (Schauf et al 1993) and releasing (Sanchez et al 1994, Surcel et al 1994) interleukin 4 (IL-4). T cells in such blood also express subnormal levels of IL-2 (Schauf et al 1993), and there is free IL-10 (N. Beyers, personal communication 1997) and tuberculosis-specific IgE antibody (Yong et al 1989). It is interesting that blood flow in skin test sites correlates with specific IgE (Gibbs et al 1991) because decreased blood flow in such sites is also a correlate of the Koch phenomenon, the tissue-damaging immunopathology that accompanies progressive disease (Anderson 1891).

Barnes and colleagues have argued that they found little Th2 activity in pleural effusions associated with tuberculosis, or in tuberculous lymphadenitis (Barnes et al 1993, Lin et al 1996, Zhang et al 1995). However, there are several objections to these conclusions. First, these are the high resistance forms of the disease, whereas immunocytochemical analysis of lesions from pulmonary tuberculosis in mice (Hernandez-Pando et al 1997) shows that the abundant Th2 cytokine-secreting cells (Figs 1b, 2b & 3b) appear in progressive disease. The same is likely to be true in humans. Second, Th2 cytokines such as IL-4 tend to be produced at levels that are at least 1000-fold lower that the levels of y-interferon (IFN-y), so a low but detectable IL-4 output may be biologically significant. In our own laboratory we have used flow cytometry of peripheral blood T cells activated in vitro with TPA (12-0-tetradecanoylphorbol 13-acetate) and ionomycin. These studies show an excess of T cells, which can be triggered to secrete IL-4, and reduced secretion of IL-2 (N. Thapa, G. Rook & J. L. Stanford, unpublished work 1997), in agreement with the mRNA studies of Schauf et al (1993). It is important to remember that immunity to M. tuberculosis seems to be sensitive to the presence of even a minor Th2 component (Hernandez-Pando et al 1997, Lindblad et al 1997).

The inappropriate Th2 component and the toxicity of tumour necrosisfactor a

Tumour necrosis factor a (TNF-a) is usually required for protection against M. tuberculosis (Flynn et al 1995a). However, this cytokine becomes toxic when there is a mycobacterial lesion with a mixed Th1+Th2 cytokine profile (Figs 1c, 2c & 3c; Hernandez-Pando et al 1997, Hernandez-Pando & Rook 1994, Rook & Hernandez-Pando 1996). This is similar to the findings in the schistosomiasis model, where immunopathology correlates with the simultaneous presence of Th1 and Th2 cytokines, and TNF-a. This immunopathology, and the permanent fibrosis and tissue damage that result, can be reduced by blocking induction of the

Th2 component (Wynn et al 1995). This phenomenon may partially explain the detrimental effect of a Th2 component in tuberculosis (Hernandez-Pando et al 1997, Lindblad et al 1997). We therefore hypothesize that the Koch phenomenon is a consequence of simultaneous release of Th1 and Th2 cytokines, and TNF-a (Koch 1891).

0 0

Post a comment