Annu. Rev. Microbiol. 1996. 50:25984 Copyright c 1996 by Annual Reviews Inc.

All rights reserved

THE PATHOGENESIS OF TUBERCULOSIS

G.A.W. Rook
Department of Bacteriology, University College London Medical School, 67-73 Riding House Street, London W1P 7LD, England

Rogelio Hernandez-Pando
Department of Pathology, Instituto Nacional de la Nutricion, Salvador Zubiran, Calle Vasco de Quiroga 15, Delegacion Tlalpan, 14000 Mexico DF
KEY WORDS: cytokine proles, immunopathology, adrenal, tumor necrosis factor

ABSTRACT
Tuberculosis patients relapse if treatment is not continued for 6 months, because chemotherapy fails to convert the patients response from the necrotizing pattern characteristic of disease (Koch phenomenon) to the nonnecrotizing bactericidal function required for optimal immunity. We need to understand the nature of these two immunological states and how to convert one to the other. Studies in mice and humans implicate differences in cytokine proles and in metabolism of adrenal steroids. Either enhanced susceptibility or protection can be evoked in mice with appropriate doses of a killed environmental saprophyte. This emphasizes the importance of shared epitopes and may explain the geographically variable efcacy of Mycobacterium bovis Bacillus Calmette Gu erin vaccination. Unlike soluble antigens of M. tuberculosis itself, which tend to evoke necrosis, the shared mycobacterial epitopes evoke little skin-test reactivity in patients. Preparations of these epitopes show potential as immunotherapeutic agents to convert the response from necrotic to bactericidal mode.

CONTENTS
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Current World Situation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Problem of the Six-Month Treatment Regimen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MECHANISM OF PROTECTIVE IMMUNITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Type 1 Response . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 260 260 260 261 261

259 0066-4227/96/1001-0259$08.00

260

ROOK & HERNANDEZ-PANDO

262 263 264 264 264 265 265 266 266 266 267 267 268 268 269 271 272 272 273 273 274 275 275 277 277 278 278

MHC Class 1Restricted Cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . T Cell Subsets of Unknown Protective Role . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nitric Oxide and the Antimycobacterial Pathways of Macrophages . . . . . . . . . . . . . . . . . TNF as a Protective Cytokine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Avoidance of Macrophage-Mediated Killing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Role of Antibody . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Pattern of Response Associated with Immunity in Human Beings . . . . . . . . . . . . . . . . MECHANISMS OF IMMUNOPATHOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Toxicity of M. tuberculosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Koch Phenomenon: The Pattern of Response Characteristic of Disease . . . . . . . . . . The Cytokine-Sensitivity of Mycobacterial Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mixed Patterns of Cytokine Expression in Tuberculosis . . . . . . . . . . . . . . . . . . . . . . . . . . . Mixed Patterns of Cytokine Expression and Tissue Damage . . . . . . . . . . . . . . . . . . . . . . . TNF -Mediated Toxicity During Murine Tuberculosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Effect of Priming a Mixed Th1 + Th2 or Th0 Response Pattern Before Infection . . . A Hypothesis to Explain the Koch Phenomenon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ANTIGENS INVOLVED IN PROTECTION AND IMMUNOPATHOLOGY . . . . . . . . . . . . . The Role of Epitopes Shared by All Mycobacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Environmental Mycobacteria and Regulation of the T Cell Response to Pathogens . . . . . Loss of Skin-Test Positivity to the Common Epitopes During Active Disease . . . . . . . . . . Immunotherapy of Tuberculosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . FACTORS THAT DEREGULATE THE TYPE 1 RESPONSE . . . . . . . . . . . . . . . . . . . . . . . . Vitamin D3 Metabolism in Tuberculous Lesions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Glucocorticoids in Tuberculosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Peripheral Regulation of Cortisol-Mediated Effects . . . . . . . . . . . . . . . . . . . . . . . . . . Adrenal Function in Human Tuberculosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

INTRODUCTION The Current World Situation

The World Health Organization has rightly declared tuberculosis to be a global emergency. One third of the worlds population is infected. Only about 5% of those infected develop active disease during the rst few years following exposure, but this represents eight million new cases each year and three million deaths. Moreover, these numbers are increasing. The stresses of poverty, malnutrition, and war increase the rate of reactivation. Even in developed countries such as the United Kingdom, the disease distribution in large cities is again identical to the distribution of poverty (102). The breakdown of healthcare systems is leading to incomplete case and contact tracing, incomplete treatment, and increases in drug resistance. In many parts of the world, most of the drugs distributed to patients are fake or out of date (54). Moreover, tuberculosis is one of the rst secondary infections to be activated in HIVpositive individuals (104).

The Problem of the Six-Month Treatment Regimen

An important reason for the current failure to control tuberculosis is that even when the best available chemotherapy is used, treatment must be continued for

PATHOGENESIS OF TUBERCULOSIS

261

at least 6 months. This treatment regimen is not a realistic proposition in most developing countries, or even in the inner cities of developed ones, because the patients feel well after a few weeks and stop taking the drugs. The solution is directly observed therapy (DOT) in which the patient is supervised while taking every dose. There are two interrelated reasons for the required six-month regimens. The rst is obvious and often discussed: The chemotherapy kills the vast majority of the bacteria within a few days, but persisters that are presumably not metabolizing (44) are not killed by the drugs. These persisters may be in true stationary phase (96), or they may be merely replicating extremely slowly in old lesions or at sites of brosis or calcication where oxygen availability may be low. The other reason for the need for prolonged treatment is usually overlooked but is, in our opinion, fundamental to an understanding of tuberculosis. Tuberculosis patients have a necrotizing pattern of response to Mycobacterium tuberculosis that is analogous to the phenomenon rst noted by Koch in guinea pigs (61). The Koch phenomenon is undeniably not a correlate of optimal protective immunity to tuberculosis. Indeed, preimmunization of animals, so that they demonstrate the Koch phenomenon before they are challenged with virulent M. tuberculosis, results in a clear and reproducible increase in susceptibility to the disease, compared to nonimmunized control animals (112). This and other aspects of the Koch phenomenon are discussed in detail below. This inappropriate pattern of response does not correct itself during treatment. Therefore, if chemotherapy is stopped at 3 months, relapse rates approach 20% (3)even in cases where treatment was an optimal rifampicin-containing chemotherapy that achieved sputum negativity well before 3 months, and in spite of the fact that there are very few live organisms in the patients tissues at this time. The task for the immunologist is therefore to understand the differences between protective immunity and the Koch phenomenon, and the factors that determine which response pattern is present. The ultimate objective is to learn how to replace the pathological response with the protective one very early in treatment. With such knowledge, we should be able to devise ultra-shortcourse chemotherapy regimens, supplemented with immunotherapy, that would provide realistic tuberculosis control in the developing world and elsewhere.

MECHANISM OF PROTECTIVE IMMUNITY The Type 1 Response

Manipulation of the immune system of mice with neutralizing antibodies or gene knockout has provided strong evidence that, in mice, immunity correlates with a type 1 response. The term type 1 is used in preference to Th1 when

262

ROOK & HERNANDEZ-PANDO

it is intended to refer to the overall pattern of cytokine release by all cell types in the infected site, rather than only that produced by the CD4+ helper T cells that were included in the original scheme of Mosmann (71). In vivo Th1 or Th2 cells act in concert with CD8+ cells and with numerous other cell types including macrophages, B cells, and some stromal cells. Collectively, these interactions give rise to two patterns of cytokine release that are known as type 1 (dominated by IL-2, IL-12, and IFN ) and type 2 (dominated by IL-4, 5, 6, 10, and 13) (20, 90). Disruption of the MHC class 2 genes or of the gene for the chain of the / T cell receptor (63), resulting in a deciency of CD4+ / T cells, renders mice susceptible even to the avirulent Mycobacterium bovis BCG. Disruption of the gene for IFN- makes mice very susceptible to M. tuberculosis (death within 3 weeks), and such mice may even die after many weeks if challenged with BCG (23, 25, 38). Similarly, supplements of IL-12 can provide some protection, though the effects are small, perhaps because in the early phase of infection mice spontaneously produce a type 1 response to M. tuberculosis (41, 24). Future experiments with gene knockout mice with nonfunctional IL-12 genes will be more informative.

MHC Class 1Restricted Cells

Mice with defective -2 microglobulin genes, rendered unable to express normal quantities of class 1 MHC products, show only a trivial increase in susceptibility to BCG (63) but are susceptible to M. tuberculosis (40). This may imply a role for CD8+ cytotoxic T cells, though proof is lacking. Such cytotoxicity could release organisms from macrophages that were failing to kill them and enable uptake of the organisms by fresh activated cells. Circulating cytotoxic T cells exist that can kill autologous macrophages infected or pulsed with antigen, but these are usually CD4+ cells (62, 56, 77). However, researchers have reported that M. tuberculosis can escape from the phagosome; this would enable antigens to gain access to the antigen processing pathway that leads to antigen presentation by MHC class 1 (73, 68). Interestingly, the more recent report suggests that the ability to escape may be a property conned to virulent strains of M. tuberculosis and not present in H37Ra or BCG, which would explain the limited effect of -2microglobulin knockout on the susceptibility of mice to BCG (63). Other authors remain unconvinced that M. tuberculosis leaves the phagosome at all (114). Other explanations exist for the effects of disrupting the gene for -2 microglobulin. This gene associates not only with class 1 MHC, but also with CD1, which, as explained below, is able to present a number of mycobacterial antigens to CD4 CD8 (double negative) T cells (7).

PATHOGENESIS OF TUBERCULOSIS

263

T Cell Subsets of Unknown Protective Role

A large proportion of human peripheral blood / T cells will proliferate in response to mycobacteria (55). Many of the responding cells express the V 2/V 2 germline genes but show diverse junctional sequences (79). In vitro, / T cells secrete a pattern of cytokines similar to the secretion pattern of Th1 cells and are cytotoxic (72, 42). Circumstantial evidence supports early involvement of / T cells in establishing immunity to tuberculosis (5). When infected with BCG, mice with disrupted genes show diminished production of IFN by spleen cells, but they do not show the increase in susceptibility seen in T cell receptor chain knockout mice (64). Thus the function of these cells remains obscure and may be partly regulatory (59). However, these experiments need to be repeated with virulent M. tuberculosis itself, because, as was shown in mice unable to express -2 microglobulin, effects may be seen when using the pathogen that are not apparent with BCG (63, 40). Preliminary evidence shows that disruption of the chain gene does indeed make mice more susceptible to M. tuberculosis (107). Considerable effort has gone into identifying the bacterial components recognized by these T cells. There is evidence for recognition of compounds that contain 5 triphosphorylated thymidine (22), and also for recognition of isopentenyl pyrophosphate and related prenyl pyrophosphate derivatives (107). The study of the / T cell response in mycobacterial disease has therefore focused attention on the fact that T cells do not recognize peptide epitopes alone.
/ T CELLS CD1 RESTRICTED RECOGNITION OF MYCOBACTERIAL LIPIDS AND GLYCOLIPIDS

Another subset of T lymphocytes also recognizes nonpeptide mycobacterial components. These are CD4 CD8 (double negative) cells that recognize antigens in association with CD1. The CD1 proteins are expressed on dendritic cells, mantle zone B cells, and cytokine-activated monocytes. They show distant homology with MHC molecules, and like MHC class 1 they associate with -2 microglobulin. However, CD1 molecules are not encoded within the MHC, and unlike the MHC they are not polymorphic. CD4 and CD8, which are involved in the interactions with MHC class 1 and class 2 respectively, are not needed for interaction with CD1. Beckmann et al have demonstrated recognition of mycolic acids by CD4 CD8 , which are / T cell receptor-bearing lymphocytes (7). Mycolic acids are -branched, -hydroxy, long-chain fatty acids that make up the bulk of the mycobacterial cell wall. Similarly, a CD4 CD8 T cell line grown from the skin lesion of a leprosy patient recognized mycobacterial lipoarabinomannan (LAM). The recognition required mannosides with (12) linkages, and the phosphatidylinositol moiety (97). Although the role of these cells is unknown,

264

ROOK & HERNANDEZ-PANDO

they appear to be cytotoxic and to secrete the Th1 cytokine pattern, which suggests that they contribute to immunity.

Nitric Oxide and the Antimycobacterial Pathways of Macrophages

Ultimately, most bacteria are probably killed by macrophages activated by the various IFN -secreting cell types described above. A relevant effector pathway in mice is dependent on nitric oxide (NO). Thus, inhibitors of NO production aggravate tuberculosis infection (as assessed by mortality, bacterial burden, and histopathology) (18, 17). The mechanism of action of the NO is uncertain: NO has important signaling and second messenger functions that may be as important as direct toxicity to the organisms (83). In the mouse, a possible major role of the IFN is activation of the inducible NO synthase. Also relevant is the Bcg phenotype because it affects the rate of replication of mycobacteria such as BCG during the early stages of infection (111). The product of this gene (Nramp-1) affects macrophage activation and may be involved in NO pathways. Interestingly, the Bcgr and Bcgs alleles may be differentially regulated by glucocorticoids (13), which is relevant to points made below about adrenal function in tuberculosis.

TNF as a Protective Cytokine

In addition to type 1 cytokines, TNF is also essential for immunity to M. tuberculosis in mice. Treatment of mice with neutralizing anti-TNF antibodies has led to dissemination of BCG infection (60), and treatment with neutralizing antibodies or knockout of the 55-kDa TNF receptor has led to rapid death from M. tuberculosis infection (39). These results further suggest a role for NO because TNF is an important trigger of NO release from IFN -activated cells, and because mycobacteria contain multiple inducers of TNF release, as do most microorganisms (2, 70). However, other effects of this cytokine may be important, and, as described below, it probably also plays a role in the immunopathology of tuberculosis.

Avoidance of Macrophage-Mediated Killing

Mycobacteria use various strategies to avoid being killed by phagocytes (16). They inhibit acidication of the phagosome (104), modify intracellular trafcking of vesicles (114), and cause quantities of LAM to insert into glycosylphosphatidylinositol (GPI)-rich domains in the cell membrane (53). LAM is itself a GPI of unusual glycan structure that has the ability to modify numerous macrophage functions, including the response to IFN , and the ability to present antigen (53). Inhibition of antigen presentation by LAM may be relevant to

PATHOGENESIS OF TUBERCULOSIS

265

the apparent inability of long-term mycobacterium-infected macrophages to present antigen to CD4+ T cells (80).

The Role of Antibody

Almost all the current review literature on the mechanism of immunity to tuberculosis states that antibody plays no role. We have found no evidence for this statement. Although antibody alone is certainly not sufcient, it may well be necessary. Spahlinger reported treatment of human tuberculosis with antisera raised by immunizing horses with M. tuberculosis that had been stored for prolonged periods, because he felt that stressed organisms were better immunogens. His results remain provocative (66, 101) and deserve some thought. A number of mycobacterial components have potent pharmacological effects. These include LAM, a variety of cytokine-inducing molecules, and inositol phosphoglycan-like (IPG) second messengers that were recently discovered in mycobacteria (GAW Rook, H Caro, E Filley & TW Rademacher, submitted for publication). Neutralization of some of these by appropriate antibody may play a role in immunity.

The Pattern of Response Associated with Immunity in Human Beings

Direct evidence for the nature of protective immunity in human beings is difcult to achieve. We can, however, deduce that protection is associated with a type 1 response by comparing patients with exposed but healthy controls (99). Similarly useful data have come from studies of cytokine expression in tuberculous lesions. Tuberculous pleuritis provides ideal clinical material, and in this high-resistance form of disease the cytokine pattern released by cell populations is heavily biased toward type 1 (4, 6). There is convincing evidence that human macrophages can be induced to release large microbicidal quantities of NO if exposed sequentially to Th2 and Th1 cytokines, or triggered with antibody that cross-links the membrane CD23 (111a). In contrast most workers are unable to achieve this result by exposing human macrophages to IFN and TNF , which is the optimal stimulus for murine macrophages. Therefore the signicance of this effector mechanism in the human disease is unclear. Workers unaware that M. tuberculosis has an active nitrate reductase have added to the confusion, because this makes the assay of nitrite as a surrogate for NO meaningless unless the medium is strictly nitrate free. Nevertheless, Hirsch et al have produced evidence that TNF can trigger inhibition of the growth of M. tuberculosis by human alveolar macrophages (50). Most attempts to make human macrophages inhibit M. tuberculosis have been unsuccessful, although this is easily achieved with murine cells by incubating them with IFN (87).

266

ROOK & HERNANDEZ-PANDO

MECHANISMS OF IMMUNOPATHOLOGY The Toxicity of M. tuberculosis

Live M. tuberculosis is inherently toxic to cells. For instance, human or murine macrophages that ingest more than about ve organisms usually die, whereas M. avium strains can multiply to remarkable numbers within cells without killing them (GAW Rook, unpublished observations). The reason for this is unknown. However, we have noted that M. tuberculosis releases a factor that greatly increases the sensitivity of infected cells to the toxicity of TNF (32, 33). In vitro M. tuberculosis readily infects nonphagocytic cells such as broblasts, and concentrations of TNF that would be growth-promoting for normal human broblasts will kill these infected cells (32). However, in the absence of TNF , M. tuberculosis is less toxic for broblasts (which do not themselves release this cytokine) than it is for macrophages, suggesting that the ability of macrophages to release TNF in the presence of mycobacterial components may be a doubleedged sword that sometimes activates successful antimycobacterial activity and sometimes kills the host cell. Although M. tuberculosis clearly has some inherent toxicity, this does not fully explain the pathology of the disease. Tuberculin and puried protein derivative (PPD) are remarkably nontoxic both in vivo and in vitro, but in suitably prepared humans or animals they provoke necrosis that is clearly due to immunopathology.

The Koch Phenomenon: The Pattern of Response Characteristic of Disease

As outlined above, Koch noted that 46 weeks after establishment of infection in guinea pigs, intradermal challenge with whole organisms or culture ltrate resulted in necrosis locally and in the original tuberculous lesion (61). A similar phenomenon occurs in humans. The tuberculin test site frequently becomes necrotic in subjects who are or have been tuberculous. Necrosis is not an inevitable consequence of the delayed hypersensitivity response to tuberculin: Necrosis does not occur when positive skin-tests to tuberculin are elicited in normal BCG recipients or in tuberculoid leprosy patients. Koch sought to exploit this phenomenon for the treatment of tuberculosis and found that injection of larger quantities of culture ltrate (Old Tuberculin) subcutaneously into tuberculosis patients would evoke necrosis in established tuberculous lesions at distant sites (1). This resulted in necrosis, sloughing, and cure of the lesions of skin tuberculosis (Lupus vulgaris, usually caused by bovine strains). However, when similar necrosis was evoked in deep lesions in the spine or lungs, the results were disastrous, and such treatment merely provided further necrotic tissue in which the bacteria could proliferate. This treatment was therefore abandoned.

PATHOGENESIS OF TUBERCULOSIS

267

The error in Kochs thinking became apparent in the 1940s. When guinea pigs were preimmunized so that they had a powerful Koch phenomenon in response to a small dose of tuberculin, they became more susceptible to tuberculosis than were nonimmunized control animals (112). Obviously, such a response was seen only if the challenge infection was introduced into the lungs or by deep intramuscular injection so that necrosis could not result in shedding of the infected tissue.

The Cytokine-Sensitivity of Mycobacterial Lesions

The ability of soluble bacterial material injected in one site to trigger necrosis in a distant tuberculous site has some parallels with the Shwartzman reaction (95). For instance, injections of endotoxin-rich material into a distant site (instead of the tuberculin used by Koch) will also trigger necrosis in tuberculous lesions (10, 28, 94), and injections into the ank of another cytokine trigger, muramyl dipeptide (MDP), causes necrosis in sites of inammation due to complete Freunds adjuvant (113). These observations are compatible with the view that tuberculous lesions are susceptible to superimposed cytokine-mediated damage. Two questions remain: Is there any evidence that TNF , in spite of its protective role, is also involved in the immunopathology of tuberculosis? If so, under what conditions do cytokines cause such damage?

Mixed Patterns of Cytokine Expression in Tuberculosis

Many symptoms of tuberculosis, such as fever, weight loss, and tissue damage, resemble the pathological effects of TNF . Evidence that these symptoms may indeed depend on TNF in human tuberculosis has come from experiments using thalidomide, which decreases the half-life of the mRNA for this cytokine (69). Patients treated with thalidomide show rapid symptomatic relief and weight gain (57). We are therefore faced with a paradox: TNF is essential for immunity but may also be responsible for pathology. Recent work provides a probable resolution of this dilemma, as described below. Although immunity to tuberculosis requires a type 1 response, in tuberculous mice type 2 cytokines are also expressed (78). The same is true in human beings, though less strikingly evident, and is most apparent when peripheral responses are considered. Studies of peripheral blood mononuclear cells precultured with mycobacterial antigen in vitro are unreliable, because the rapid production of IFN from NK cells and from the Th1 cells that are almost always present tends to inhibit full expression of Th2 cytokines. Nevertheless, both Th1 and Th2 cells are seen by such methods (105). If peripheral blood lymphocytes are examined without preculture, it becomes apparent that the IL-4 gene is indeed expressed in patients peripheral blood mononuclear cells (92), while there is a decit in IL-2 expression (92). This suggests the presence of a Th2

268

ROOK & HERNANDEZ-PANDO

component in disease, supported by the fact that tuberculosis patients have IgE antibody (115) and IgG4 antibody (GAW Rook, unpublished observations). Both of these antibodies are IL-4-dependent in human beings. When using a laser Doppler velicometer to study tuberculin test sites, Gibbs et al found that the extent to which blood ow was reduced in the center of the site at both 6 h and 48 h was related to the level of specic IgE antibody to M. tuberculosis (43). This appears to indicate incipient necrosis.

Mixed Patterns of Cytokine Expression and Tissue Damage

The above points are of considerable importance because the combination of Th1 + Th2 + TNF results in severe pathological damage in other systems. For instance, the granulomata formed in response to the ova of Schistosoma in mice (45) depend on the simultaneous presence of all three elements. These granulomata, like mycobacterial lesions (10, 28, 94, 74), are acutely sensitive to further tissue damage if systemic cytokine release is induced (15, 31). Moreover, if the Th2 component is reduced by preimmunization with ovum antigens plus IL-12, the granulomata are much smaller, and, of great signicance, the residual tissue damage and brosis are reduced (113). A correlation is readily demonstrated between the cytokine prole of the helper T cells that are involved in an inammatory site and the sensitivity of that site to TNF . When mice were immunized with a low dose (85) of an intensely immunogenic killed mycobacterial preparation (autoclaved Mycobacterium vaccae), only Th1 cytokine production was primed (75). If as much as 1 g of TNF was injected into delayed hypersensitivity (DTH) response sites that were elicited 24 h earlier in such animals, no necrosis was caused. However, the 100-fold larger dose (109 ) of the same killed M. vaccae preparation evoked a mixed pattern with priming for both Th1 and Th2 cytokine secretion. Injection of TNF into DTH response sites that were elicited in these animals resulted in necrosis (75). Therefore, TNF released into a relatively pure Th1-mediated inammatory site may act as a mere supplementary macrophage-activating molecule, but when released into a mixed Th1 + Th2 or possibly Th0 site, it causes damage. Therefore, the interrelatedness of decreased blood ow in tuberculin test sites and the level of specic IgE antibody to M. tuberculosis may be relevant (43). In fact, these results must still be interpreted with caution because we do not know what correlate of the Th2 response pattern is responsible for the increased tissue damage.

TNF -Mediated Toxicity During Murine Tuberculosis

Conrmation of the relevance of the observations outlined above has emerged from a study of the TNF -sensitivity of DTH response sites elicited in mice with pulmonary tuberculosis (47). During the rst 3 weeks, i.e. the period of

PATHOGENESIS OF TUBERCULOSIS

269

Figure 1 The effect of 1 g of recombinant murine TNF on the swelling of delayed hypersensitivity reactions at different stages of pulmonary tuberculosis in Balb/c mice. Foot-pads were challenged with 20 g of M. tuberculosis antigen, and the swelling was determined at 24 h (square). Then 1 g of TNF was injected into the same site, and swelling was re-assessed 20 h later (44 h) (lled circle). Means standard deviation (SD) are shown. Without TNF , all reactions decline by 44 h (not shown). From reference 47, with permission.

type 1 response (78), DTH sites were not sensitive to TNF . After 50 days, the animals enter a phase of slowly progressing disease that is accompanied by Th2 cytokine production and high IgG1 antibody titres (a Th2-associated murine subclass). In these animals, DTH sites become TNF -sensitive (Figure 1). At this time the adrenals undergo atrophy (Figure 2), which, by compromising glucocorticoid feedback, may further aggravate toxicity of TNF (47). This theme is developed below.

The Effect of Priming a Mixed Th1 + Th2 or Th0 Response Pattern Before Infection
The argument that the mixed Th1 + Th2 cytokine pattern is associated with pathology is further strengthened by looking at the consequences of generating in normal mice the immunological state seen at day 50 in infected mice. When mice were preimmunized with 107 M. vaccae [the optimal dose for inducing a Th1 non-TNF -sensitive response (48)], they were partially protected (Figure 3). In sharp contrast, when mice were preimmunized using the dose of 109 killed M. vaccae, they developed the mixed Th1 + Th2 response pattern with TNF -sensitive DTH responsiveness described above (48) and

270

ROOK & HERNANDEZ-PANDO

Figure 2 The weights of the right adrenals of male Balb/c mice infected into the trachea with 106 M. tuberculosis. Means SD of groups of 35 mice are shown. Points marked with an asterisk differ from the control value (p < 0.025, Students t test). From reference 47, with permission.

Figure 3 The effect of priming with 1 107 (triangle) or 1 109 (lled circle) autoclaved M. vaccae or saline (square) 2 months before intratracheal infection on the survival of Balb/c mice. Depending on the immunization protocol, the common epitopes can mediate responses causing either protection (triangle) or increased susceptibility (lled circle). Asterisks indicate results that are signicantly different from saline controls by Fishers exact test.

PATHOGENESIS OF TUBERCULOSIS

271

Figure 4 The effect of priming with 1107 (triangle) or 1109 (lled circle) autoclaved M. vaccae or saline (square) on changes in left adrenal weight induced by intratracheal infection with M. tuberculosis (1106 ) 2 months later. Means of 35 mice SD. Adapted from reference 47 with permission.

were found to be more susceptible to intratracheal M. tuberculosis than were unimmunized control animals (48) (Figure 3). Moreover, their adrenals atrophied within seven days of infection (48) (Figure 4), further illustrating the parallels between the immunological state evoked by 109 killed M. vaccae and the state that accompanies late progressive disease.

A Hypothesis to Explain the Koch Phenomenon

Signicantly, the Koch phenomenon accompanies progressive disease in guinea pigs and human beings, so, it is logical to ask whether the immunological state evoked by 109 M. vaccae in mice is equivalent to the Koch phenomenon in these species. We have already pointed out that the Koch phenomenon increases susceptibility to the disease in guinea pigs (112), just as the state evoked by 109 M. vaccae does in mice. We put this forward as a hypothesis, but we emphasize that these results do not tell us which correlate of the Th2 response leads to toxicity of TNF in mixed inammatory sites or to the increased susceptibility to disease. The relevant factor may not be one of the conventional Th2 cytokines but rather some unidentied factor that tends to be activated by protocols that evoke Th2 T cell responses. This important point can now be tested by experimentation.

272

ROOK & HERNANDEZ-PANDO

ANTIGENS INVOLVED IN PROTECTION AND IMMUNOPATHOLOGY The Role of Epitopes Shared by All Mycobacteria
As stated above, epitopes common to M. tuberculosis and to an environmental saprophyte (M. vaccae) are capable of acting as targets for a protective type 1 response or for a detrimental TNF -sensitive, mixed Th1 + Th2 or Th0 response. This should not surprise immunologists, but it often does. We have known for many years that the Bacillus Calmette-Gu erin (BCG) is as effective a vaccine against leprosy as it is against tuberculosis (35). Therefore, BCG must be able to work through common epitopes (Figure 5). Similarly, Fine et al reviewed evidence that contact with an environmental organism that leads to mycobacterial skin-test positivity is protecting the population of Malawi from both tuberculosis and leprosy (36). Also, Silva & Lowrie found that the 65kDa heat shock protein of M. leprae can protect mice against M. tuberculosis (98), and the importance to protective immunity of conserved proteins such as heat shock proteins has been emphasized repeatedly by other researchers (116). The 30-kDa proteins are also strongly conserved within the genus and can vaccinate guinea pigs (52). In spite of these ndings, researchers hold a deep prejudice against the view that protection can be mediated via epitopes that

Figure 5 A diagrammatic representation of the antigenic relationship between environmental saprophytes and M. tuberculosis. The antigens in the shaded area may include conserved proteins such as heat shock proteins. They can mediate protection but not the Koch phenomenon, so they appear to be handled differently by the immune system. Possible explanations are discussed in the text.

PATHOGENESIS OF TUBERCULOSIS

273

are not species specic. This prejudice dates from the era of the early antibodymediated vaccines, in which antibodies neutralize microbial components by binding conformational epitopes on toxins, enzymes, or adhesion molecules. These substances are often species restricted. The fact that T cells do not neutralize anything but instead recognize short peptide sequences cleaved from microbial proteins, together with the fact that T cells are not taxonomists, should be sufcient to dispel such a prejudice. The concept of species-specicity is irrelevant to T cell function.

Environmental Mycobacteria and Regulation of the T Cell Response to Pathogens

Because mycobacteria are not part of the normal commensal ora of human beings, the nature, route, and dose of mycobacterial contact are variables that depend on where and how an individual lives. This variable priming of antimycobacterial responses by saprophytes, which, as explained above, can either protect from or predispose to infection, almost certainly explains the variable efcacy of BCG in different parts of the world (34) and the variability of the protective effect of saprophytic species in some environments (36).

Loss of Skin-Test Positivity to the Common Epitopes During Active Disease

Tuberculosis patients who still maintain necrotizing skin-test positivity to antigens of M. tuberculosis itself have diminished or absent skin-test responses to environmental saprophytes (58). This remarkable paradox implies that the common epitopes are handled in quite a different way from the tuberculosis-specic components. At least three hypotheses may be proposed for the selective loss of skin-test responsiveness to common epitopes. First, the common epitopes are by denition common and are encountered from birth in a variety of mycobacteria (and related genera). Most of this exposure is by the oral route. Therefore, as with other antigens encountered by this route, there has been possible priming of regulatory TGF -producing T cells (19). According to a rst hypothesis, these regulatory cells may switch off the response to the common epitopes during the infection. A second hypothesis is that because the common epitopes have been encountered at low dose over many years, the response to them has become locked into the Th1 mode and cannot be converted to Th2 or Th0 (11). A third possibility is that the common epitopes are handled in an intrinsically different manner by the immune response because they are situated in highly conserved proteins such as heat shock proteins that are potential targets of autoimmunity (21).

274

ROOK & HERNANDEZ-PANDO

Figure 6 The effect of a single immunotherapeutic injection of 107 (triangle) or 109 (lled circle) autoclaved M. vaccae or saline (square), 60 days after intratracheal infection of Balb/c mice. Only the Th1-inducing dose (triangle) is benecial. The dose that evokes a mixed Th1 + Th2 pattern, with increased sensitivity to TNF , exacerbates the disease (lled circle). Asterisks indicate results that are signicantly different from saline controls by Fishers exact test.

Immunotherapy of Tuberculosis
The common epitopes are capable of initiating protective responses, so the loss of responsiveness to them in the disease state (no matter what the reason) suggests the possibility that they can be used therapeutically. Crude preparations of M. tuberculosis itself cannot be used because, as Robert Koch found to his cost, they evoke necrosis (1). However, a killed preparation of an environmental saprophyte that has the appropriate Th1-adjuvant capability theoretically may be able to restore Th1 and DTH responses to the common epitopes, while evoking no necrosis. An autoclaved preparation of M. vaccae contains the common epitopes and has suitable adjuvant properties (48). When used at the optimal Th1-inducing dose (85), such a preparation exerts signicant benecial effects in a murine model of pulmonary tuberculosis when given on day 60, during the late mixed Th1 + Th2 or Th0 phase of the disease (Figure 6; S Baldwin & I Orme, unpublished observations). These effects are equivalent to immunotherapy of multi-drug-resistant disease in human beings in light of the fact that no chemotherapy was given. After encouraging pilot studies in human beings (30, 76, 103), this material is now undergoing Phase 3 efcacy trials in

PATHOGENESIS OF TUBERCULOSIS

275

Durban, Republic of South Africa, in collaboration with Dr. Bernard Fourie of the South African Medical Research Council. The trial will be decoded in late 1996. An alternative approach, also based on the need to boost type 1 responses, is the administration of cytokines. IFN may be effective in the rare nonHIVinfected individuals infected with M. avium (51) because these patients often have a decit in production of this cytokine. However, therapy with IFN looks less promising in tuberculosis. Tuberculosis patients have decient release of IL-2 from peripheral blood lymphocytes (92); therefore, attempts are now in progress to treat tuberculosis with this cytokine (G Kaplan, personal communication).

FACTORS THAT DEREGULATE THE TYPE 1 RESPONSE

We have argued above that an inappropriate Th2 component is present in both murine and human tuberculosis. Its presence becomes more striking in human tuberculosis as the disease becomes more severe. What then are the likely causes of this shift in cytokine prole? Increasing antigen load is likely to be one factor, owing to the striking linkage of the Th1 + Th2 balance to dose when immunizing with particulate antigens such as mycobacteria (48) or leishmania (11). Similarly, prostaglandin release may play a part (109, 49). However, strong reasons now exist for suggesting that endocrine interactions with the immune system are important to the changing cytokine prole.

Vitamin D3 Metabolism in Tuberculous Lesions

The macrophages of tuberculosis patients, following activation by IFN- , express an active 1 -hydroxylase and rapidly convert 25(OH)-vitamin-D3 to calcitriol (88, 84) (Figure 7). Their T cells may also express this enzyme (14). This potent phenomenon leads occasionally to leakage of calcitriol into the periphery and to hypercalcemia, although its role in the disease has in the past been difcult to understand (84). This may be a feedback mechanism that tends to down-regulate Th1 and enhance Th2 responses, a possibility recently suggested because the active vitamin D3 metabolite 1,25(OH)2 cholecalciferol (calcitriol) inhibits production of IFN- and IL-2 and increases production of IL-4 and IL5 (82, 27). This Th1-to-Th2 switching may well be related to the ability of calcitriol to inhibit release of IL-12 (65). The true physiological relevance of these effects in vivo remains unproven but now seems likely. The synthesis of novel analogues of calcitriol that have less tendency to cause hypercalcemia has allowed them to be tested as suppressors of Th1 responses in models in vivo. Some of these analogues are strikingly effective and will prolong allograft survival as well as reduce the requirement for cyclosporin A in treatment (110).

276

ROOK & HERNANDEZ-PANDO

Figure 7 The mechanisms controlling the functions of glucocorticoids in lymphoid tissue. Cortisol arriving from circulation may be inactivated by the stromal cells, which convert it to cortisone. Alternatively, its effects may be opposed by unidentied metabolites of DHEA, following desulphation of the circulating DHEAS by macrophages. The latter also contain 1 -hydroxylase and form calcitriol [1,25(OH)2 cholecalciferol], which may further deviate the response towards Th2.

In the 1940s, attempts were made to treat tuberculosis with vitamin D. When patients with skin tuberculosis (Lupus vulgaris, often due to M. bovis) were treated with this vitamin, the chronic nonhealing granulomatous lesions often underwent necrosis followed by resolution (67). However, necrosis and liquefaction also occurred in deep lesions in the spine and lungs (12), so the results were as disastrous as the use of Kochs immunotherapy, described above (1). A very speculative explanation would be that the additional priming of macrophages for cytokine release attributable to the calcitriol formed in the lesions (89), combined with the Th1-to-Th2 shift that this metabolite also causes, was sufcient to exacerbate the Koch phenomenon, which resulted in sloughing of skin lesions and liquefaction of deep ones.

PATHOGENESIS OF TUBERCULOSIS

277

Glucocorticoids in Tuberculosis
Adrenal steroids may also contribute to the dysfunction of Th1 responses in tuberculosis. Reactivation or progression of infection with tuberculosis is sensitive to activation of the hypothalamo-pituitary adrenal axis. Exposure of humans to the stress of war or poverty (102), or cattle to the stress of transportation, is enough to cause reactivation of disease. The disease-promoting effect of stress has been demonstrated under more controlled conditions in mice (13, 108). These effects are thought to be mediated via glucocorticoid release, for cortisol (corticosterone in mice) reduces macrophage activation and Th1 T cell activity (27) while synergizing with some Th2 functions (37). Thus the mechanisms that control tuberculosis are sensitive to glucocorticoids, probably because glucocorticoids provoke a Th1-to-Th2 shift. Several other features of tuberculosis are compatible with glucocorticoidmediated effects. These include a reduced CD4 count, a reduced CD4/CD8 ratio (85, 86, 100, 106), and a mildly impaired glucose tolerance (117). The almost total loss of the evening glucocorticoid trough indicates that the periphery is indeed exposed over a 24-h period to increased cortisol levels, even in those patients in whom early morning serum cortisol is normal (91). In contrast, other aspects of tuberculosis suggest reduced adrenal reserve. Some tuberculosis patients die suddenly and without obvious cause during treatment, and adrenal decit has often been the suspected cause (75, 93). Occasionally the adrenals are themselves infected, but there are patients whose adrenals are found in postmortem examination to be small and without evidence of direct infection, as in tuberculous mice (47). Perhaps also of signicance, inhibition of cytokine-mediated tissue damage requires rapid peaks of cortisol in response to cytokine signals to the hypothalamo-pituitary-adrenal (HPA) axis (8, 118). TNF and IL-1 are much more toxic in adrenalectomized than in control animals (118, 8). Thus, reduced adrenal reserve could play a role in the toxicity of TNF , discussed above (57), and in the toxicity of TNF in tuberculous mice once they have entered the phase of adrenal atrophy (47). The atrophic adrenals of these mice were not infected.

The Peripheral Regulation of Cortisol-Mediated Effects

The concept of increased peripheral cortisol effects in the presence of reduced adrenal reserve is compatible with what we know of the regulation of steroid synthesis, but previous studies of adrenal function in tuberculosis (reviewed in 81) were mostly conned to the investigation of adrenal reserve using vastly supraphysiological quantities of adrenocorticotrophic hormone (ACTH). Such studies failed to address the question in relation to new insights into the way in which the peripheral effects of cortisol on the T cell system are regulated

278

ROOK & HERNANDEZ-PANDO

(Figure 7). Briey, cortisol function within lymphoid tissue is regulated by local production of metabolites of dehydroepiandrosterone sulfate (DHEAS) that have antiglucocorticoid effects (9), and by conversion of cortisol into inactive cortisone by an 11 -hydroxysteroid dehydrogenase (11 OHSD) present in the stromal cells (26, 29). Both inhibition of DHEA sulfatase (R Foulkes, personal communication) and of 11 OHSD (R Daynes, personal communication) have a profound glucocorticoid-like effect. In contrast, administration of DHEA or 3,17-androstenediol causes an antiglucocorticoid effect and a Th1 bias (9, 26). Therefore, study of DHEA/cortisol and cortisone/cortisol ratios is important, as are attempts to understand any changes in the metabolism of DHEA.

Adrenal Function in Human Tuberculosis

Studies of 24-h urine samples in which adrenal steroid metabolites were identied and quantitated by gas chromatography and mass spectrometry have revealed changes that are compatible with a disturbance in the regulatory balances outlined above (85a). The total output both of cortisol derivatives and of androgens was frequently reduced by as much as 50%. Cortisol also underwent reduced conversion to cortisone and cortolones, so that tetrahydrocortisol levels in urine were normal or even raised. This reduced conversion explains the normal serum cortisol concentrations reported by others (81) and must be due either to decreased activity of 11 OHSD or to increased activity of a reductase. The site of the enzyme changes that cause this cortisone/cortisol imbalance is not yet known, but the implications for the regulatory system shown in Figure 7 are clear. Similarly, a decreased DHEA/cortisol ratio was evident, as was decreased conversion of DHEA to reduced forms (etiocholanolone and androsterone) and unchanged or increased conversion to 16 -hydroxylated forms. Because we do not know which metabolites of DHEA exert the antiglucocorticoid effects, interpretation of this nding is not yet possible, but it is suggestive. There is no reason to suppose that these changes are disease specic, but they may play a role in pathogenesis. Manipulation of these regulatory circuits has profound effects on the course of pulmonary tuberculosis in mice. For instance, administration of antiglucocorticoid steroids is protective during the early phase of murine tuberculosis that is accompanied by adrenal hypertrophy and presumably by corticosterone release (47) (see Figure 2). However, administration during the subsequent phase of adrenal atrophy (see Figure 2) can be fatal (R Hernandez-Pando & GAW Rook, manuscript in preparation).

CONCLUSIONS
Studies at the molecular level have yielded much information about the nature of protection, the difference between protection and immunopathology, the

PATHOGENESIS OF TUBERCULOSIS

279

identity of the most protective groups of antigens, and the crucial immunoendocrine interactions. We have now reached the point where we can stand back from the detail and attempt a physiological view of the entire disease process, and integrate it with an ecological view of the role of exposure to mycobacteria in our rapidly changing environment. These considerations lead to the prospect of simple novel types of clinical intervention that are much needed in the present global emergency.

Literature Cited 1. Anderson MC. 1891. On Kochs treatment. Lancet 1:65152 2. Averill L, Toossi Z, Aung H, Boom WH, Ellner JJ. 1995. Regulation of production of tumor necrosis factor alpha in monocytes stimulated by the 30kilodalton antigen of Mycobacterium tuberculosis. Infect. Immun. 63:32068 3. Balasubramanian R, Sivasubramanian S, Vijayan VK, Ramachandran R, Jawahar MS, et al. 1990. Five year results of a 3 month and two 5 month regimens for the treatment of sputum-positive pulmonary tuberculosis in South India. Tubercle 71:25358 4. Barnes PF, Fong SJ, Brennan PJ, Twomey PE, Mazumder A, Modlin RL. 1990. Local production of tumor necrosis factor and IFN-gamma in tuberculous pleuritis. J. Immunol. 145:14954 5. Barnes PF, Grisso CL, Abrams JS, Band H, Rea TH, Modlin RL. 1992. Gamma delta T lymphocytes in human tuberculosis. J. Infect. Dis. 165:50612 6. Barnes PF, Lu S, Abrams JS, Wang E, Yamamura M, Modlin RL. 1993. Cytokine production at the site of disease in tuberculosis. Infect. Immun. 61:348289 7. Beckman EM, Porcelli SA, Morita CT, Behar SM, Furlong ST, Brenner MB. 1994. Recognition of a lipid antigen by CD1-restricted alpha beta+ T cells. Nature 372:69194 8. Bertini R, Bianchi M, Ghezzi P. 1988. Adrenalectomy sensitizes mice to the lethal effects of interleukin 1 and tumor necrosis factor. J. Exp. Med. 167:1708 12 9. Blauer KL, Poth M, Rogers WM, Bernton EW. 1991. Dehydroepiandrosterone antagonises the suppressive effects of dexamethasone on lymphocyte proliferation. Endocrinology 129(6):3174 79 Bordet P. 1931. Contribution l etude de lallergie. C. R. Soc. Biol. 107:62223 Bretscher PA, Wei G, Menon JN, Bielefeldt-Ohmann H. 1992. Establishment of stable, cell-mediated immunity that makes susceptible mice resistant to Leishmania major. Science 257:53942 Brincourt J. 1967. Le calcif erol a-t-il une action liqu eante sur le caseum? Poumon Coeur 23:84151 Brown DH, LaFuse W, Zwilling BS. 1995. Cytokine-mediated activation of macrophages from Mycobacterium bovis BCG-resistant and -susceptible mice: differential effects of corticosterone on antimycobacterial activity and expression of the Bcg gene (candidate Nramp). Infect. Immun. 63:298388 Cadranel J, Garabedian M, Milleron B, Guillozo H, Akoun G, Hance AJ. 1990. 1,25(OH)2 D3 production by T lymphocytes and alveolar macrophages recovered by lavage from normocalcaemic patients with tuberculosis. J. Clin. Invest. 85:158893 Carswell EA, Old LJ, Kassel RL, Green S, Fiore W, Williamson B. 1975. An endotoxin-induced serum factor that causes necrosis of tumours. Proc. Natl. Acad. Sci. USA 72:366670 Chan J, Fan X, Hunter SW, Brennan PJ, Bloom BR. 1991. Lipoarabinomannan, a possible virulence factor involved in persistence of Mycobacterium tuberculosis within macrophages. Infect. Immun. 59:175561 Chan J, Tanaka K, Carroll D, Flynn J, Bloom BR. 1995. Effects of nitric ox-

10. 11.

12. 13.

14.

15.

16.

17.

280

ROOK & HERNANDEZ-PANDO

ide synthase inhibitors on murine infection with Mycobacterium tuberculosis. Infect. Immun. 63:73640 Chan J, Xing Y, Magliozzo RS, Bloom BR. 1992. Killing of virulent Mycobacterium tuberculosis by reactive nitrogen intermediates produced by activated murine macrophages. J. Exp. Med. 175:111122 Chen Y, Inobe J, Marks R, Gonnella P, Kuchroo VK, Weiner HL. 1995. Peripheral deletion of antigen-reactive T cells in oral tolerance. Nature 376:17780 Clerici M, Shearer GM. 1994. The Th1Th2 hypothesis of HIV infection: new insights. Immunol. Today 15:57581 Cohen IR, Young DB. 1991. Autoimmunity, microbial immunity and the immunological homunculus. Immunol. Today 12:10510 Constant P, Davodeau F, Peyrat M, Poquet Y, Puzo G, et al. 1994. Stimulation of human g/d T cells by nonpeptidic mycobacterial ligands. Science 264:26770 Cooper AM, Dalton DK, Stewart TA, Grifn JP, Russell DG, Orme IM. 1993. Disseminated tuberculosis in interferon gamma gene-disrupted mice. J. Exp. Med. 178:224347 Cooper AM, Roberts AD, Rhoades ER, Callahan JE, Getzy DM, Orme IM. 1995. The role of interleukin-12 in acquired immunity to Mycobacterium tuberculosis infection. Immunology 84:42332 Dalton DK, Pitts-Meek S, Keshav S, Figari IS, Bradley A, Stewart TA. 1993. Multiple defects of immune cell function in mice with disrupted interferongamma genes. Science 259:173942 Daynes RA, Araneo BA, Hennebold J, Enioutina E, Mu HH. 1994. Steroids as essential regulators of the mammalian immune response. J. Invest. Dermatol. 105:S15S19 Daynes RA, Meikle AW, Araneo BA. 1991. Locally active steroid hormones may facilitate compartmentalization of immunity by regulating the types of lymphokines produced by helper T cells. Res. Immunol. 142:4045 Debonera G, Tzortakis N, Falchetti E. 1932. Inammation et ph enomene de Shwartzman. C. R. Soc. Biol. 109:2426 Dougherty TF, Berliner ML, Berliner DL. 1960. 11b-hydroxy dehydrogenase system activity in thymi of mice following prolonged cortisol treatment. Endocrinology 66:55058 Etemadi A, Farid R, Stanford JL. 1992. Immunotherapy for drug-resistant tuberculosis. Lancet 340:136061 31. Ferluga J, Doenhoff MJ, Allison AC. 1979. Increased hepatotoxicity of bacterial lipopolysaccharide in mice infected with Schistosoma mansoni. Parasite Immunol. 1:28994 32. Filley EA, Bull HA, Dowd PM, Rook GAW. 1992. The effect of Mycobacterium tuberculosis on the susceptibility of human cells to the stimulatory and toxic effects of Tumour Necrosis Factor. Immunology 77:5059 33. Filley EA, Rook GAW. 1991. Effect of mycobacteria on sensitivity to the cytotoxic effects of tumor necrosis factor. Infect. Immun. 59:256772 34. Fine PEM. 1993. Immunities in and to tuberculosis: implications for pathogenesis and vaccination. In Tuberculosis: Back to the Future. Proc. London Sch. Hyg. Trop. Med. 3rd Annu. Public Health Forum, ed. KPWJ McAdam, JDH Porter, pp. 5478. Chichester, England: Wiley 35. Fine PEM, Ponnighaus JM, Maine N, Clarkson JA, Bliss L. 1986. The protective efcacy of BCG against leprosy in Northern Malawi. Lancet 2:499502 36. Fine PEM, Sterne JAC, Ponnighaus JM, Rees RJW. 1994. Delayed type hypersensitivity, mycobacterial vaccines and protective immunity. Lancet 344:1245 49 37. Fischer A, Konig W. 1991. Inuence of cytokines and cellular interactions on the glucocorticoid-induced Ig (E, G, A, M) synthesis of peripheral blood mononuclear cells. Immunology 74:22833 38. Flynn JL, Chan J, Triebold KJ, Dalton DK, Stewart TA, Bloom BR. 1993. An essential role for interferon gamma in resistance to Mycobacterium tuberculosis infection. J. Exp. Med. 178:224954 39. Flynn JL, Goldstein MM, Chan J, Triebold KJ, Pfeffer K, et al. 1995. Tumor necrosis factor-alpha is required in the protective immune response against Mycobacterium tuberculosis in mice. Immunity 2:56172 40. Flynn JL, Goldstein MM, Triebold KJ, Koller B, Bloom BR. 1992. Major histocompatibility complex class I-restricted T cells are required for resistance to Mycobacterium tuberculosis infection. Proc. Natl. Acad. Sci. USA 89:12013 17 41. Flynn JL, Goldstein MM, Triebold KJ, Sypek J, Wolf S, Bloom BR. 1995. IL-12 increases resistance of BALB/c mice to Mycobacterium tuberculosis infection.

18.

19.

20. 21.

22.

23.

24.

25.

26.

27.

28. 29.

30.

PATHOGENESIS OF TUBERCULOSIS
J. Immunol. 155:251524 42. Follows GA, Munk ME, Gatrill AJ, Conradt P, Kaufmann SH. 1992. Gamma interferon and interleukin 2, but not interleukin 4, are detectable in gamma/delta T-cell cultures after activation with bacteria. Infect. Immun. 60:122931 43. Gibbs JH, Grange JM, Beck JS, Jawad E, Potts RC, et al. 1991. Early delayed hypersensitivity responses in tuberculin skin tests after heavy occupational exposure to tuberculosis. J. Clin. Pathol. 44:91923 44. Grange JM. 1992. The mystery of the mycobacterial persister. Tubercle Lung Dis. 73:24951 45. Grzych JM, Pearce E, Cheever A, Caulada ZA, Caspar P, et al. 1991. Egg deposition is the stimulus for the production of Th2 cytokines in murine schistosomiasis mansoni. J. Immunol. 146:132240 46. Hawken M, Nunn P, Gathua S, Brindle R, Godfrey-Faussett P, et al. 1993. Increased recurrence of tuberculosis in HIV-1-infected patients in Kenya. Lancet 342:33237 47. Hernandez-Pando R, Orozco H, Honour JP, Silva J, Leyva R, Rook GAW. 1995. Adrenal changes in murine pulmonary tuberculosis; a clue to pathogenesis? FEMS Immunol. Med. Microbiol. 12:6372 48. Hernandez-Pando R, Rook GAW. 1994. The role of TNFa in T cell-mediated inammation depends on the Th1/Th2 cytokine balance. Immunology 82:59195 49. Hilkens CM, Vermeulen H, van-Neerven RJ, Snijdewint FG, Wierenga EA, Kapsenberg ML. 1995. Differential modulation of T helper type 1 (Th1) and T helper type 2 (Th2) cytokine secretion by prostaglandin E2 critically depends on interleukin-2. Eur. J. Immunol. 25:5963 50. Hirsch CS, Ellner JJ, Russell DG, Rich EA. 1994. Complement receptor mediated uptake and Tumour Necrosis Factor a-mediated growth inhibition of Mycobacterium tuberculosis by human alveolar macrophages. J. Immunol. 152:74353 51. Holland SM, Eisenstein EM, Kuhns DB, Turner ML, Fleisher TA, et al. 1994. Treatment of refractory disseminated nontuberculous mycobacterial infection with interferon gamma. A preliminary report. N. Engl. J. Med. 330:134855 52. Horwitz MA, Lee BW, Dillon BJ, Harth G. 1995. Protective immunity against tuberculosis induced by vaccination with

281

53.

54. 55.

56.

57. 58.

59.

60.

61. 62.

63.

major extracellular proteins of Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA 92:153034 Ilangumaran S, Arni S, Poincelet M, Theler JM, Brennan PJ, et al. 1995. Integration of mycobacterial lipoarabinomannans into glycosylphosphatidylinositol-rich domains of lymphomonocytic cell plasma membranes. J. Immunol. 155:133442 Ityavyar DA. 1988. Health service inequalities in Nigeria. Soc. Sci. Med. 27:122335 Kabelitz D, Bender A, Schondelmaier S, Schoel B, Kaufmann SH. 1990. A large fraction of human peripheral blood gamma/delta + T cells is activated by Mycobacterium tuberculosis but not by its 65-kD heat shock protein. J. Exp. Med. 171:66779 Kaleab B, Ottenoff T, Converse P, Halapi E, Tadesse G, et al. 1990. Mycobacterialinduced cytotoxic T cells as well as nonspecic killer cells derived from healthy individuals and leprosy patients. Eur. J. Immunol. 20:265159 Kaplan G. 1994. Cytokine regulation of disease progression in leprosy and tuberculosis. Immunobiology 191:56468 Kardjito T, Beck JS, Grange JM, Stanford JL. 1986. A comparison of the responsiveness to four new tuberculins among Indonesian patients with pulmonary tuberculosis and healthy subjects. Eur. J. Respir. Dis. 69:14245 Kaufmann SH, Blum C, Yamamoto S. 1993. Crosstalk between alpha/beta T cells and gamma/delta T cells in vivo: activation of alpha/beta T-cell responses after gamma/delta T-cell modulation with the monoclonal antibody GL3. Proc. Natl. Acad. Sci. USA 90:962024 Kindler V, Sappino AP, Grau GE, Piguet PF, Vassalli P. 1989. The inducing role of tumor necrosis factor in the development of bactericidal granulomas during BCG infection. Cell 56:73140 Koch R. 1891. Fortsetzung u ber ein Heilmittel gegen Tuberculose. Dtsch. Med. Wochenschr. 17:1012 Kumararatne DS, Drysdale P, Gaston JS, Stacey P, Richardson P, Wise R. 1988. Mycobacterium tuberculosis antigen specic human T-cell lines are cytolytic to autologous antigen pulsed macrophages. Adv. Exp. Med. Biol. 237:4016 Ladel CH, Daugelat S, Kaufmann SH. 1995. Immune response to Mycobacterium bovis bacille Calmette Guerin in-

282

ROOK & HERNANDEZ-PANDO

fection in major histocompatibility complex class I- and II-decient knock-out mice: contribution of CD4 and CD8 T cells to acquired resistance. Eur. J. Immunol. 25:37784 Ladel CH, Hess J, Daugelat S, Mombaerts P, Tonegawa S, Kaufmann SH. 1995. Contribution of alpha/beta and gamma/delta T lymphocytes to immunity against Mycobacterium bovis bacillus Calmette Guerin: studies with T cell receptor-decient mutant mice. Eur. J. Immunol. 25:83846 Lemire JM. 1994. Immunomodulatory actions of 1,25-dihydroxyvitamin D3. J. Steroid. Biochem. Mol. Biol. 53:599 602 Macassey L, Saleeby CW. 1934. Spahlinger Contra Tuberculosis 1908 1934. An International Tribute. London: Bale /Danielsson. 271 pp. Macrae DE. 1947. Calciferol treatment of Lupus vulgaris. Br. Med. J. 59:33338 McDonough KA, Kress Y, Bloom BR. 1993. Pathogenesis of tuberculosis: interaction of Mycobacterium tuberculosis with macrophages. Infect. Immun. 61:276373 Moreira AL, Sampaio EP, Zmuidzinas A, Frindt P, Smith KA, Kaplan G. 1993. Thalidomide exerts its inhibitory action on tumor necrosis factor alpha by enhancing mRNA degradation. J. Exp. Med. 177:167580 Moreno C, Taverne J, Mehlert A, Bate CA, Brealey RJ, et al. 1989. Lipoarabinomannan from Mycobacterium tuberculosis induces the production of Tumour Necrosis Factor from human and murine macrophages. Clin. Exp. Immunol. 76:24045 Mosmann TR, Cherwinski H, Bond MW, Giedlin MA, Coffman RL. 1986. Two types of murine helper T cell clone. 1. Denition according to proles of lymphokine activities and secreted proteins. J. Immunol. 136:234857 Munk ME, Gatrill AJ, Schoel B, Gulle H, Pfeffer K, et al. 1990. Immunity to mycobacteria: possible role of alpha/beta and gamma/delta T lymphocytes. Acta Pathol. Microbiol. Immunol. Scand. Sect. A 98:66973 Myrvik QN, Leake ES, Wright MJ. 1984. Disruption of phagosomal membranes of normal alveolar macrophages by the H37Rv strain of Mycobacterium tuberculosis. A correlate of virulence. Am. Rev. Respir. Dis. 129:32228 Nagao S, Tanaka A. 1985. Necrotic inammatory reaction induced by muramyl dipeptide in guinea-pigs sensitized by tubercle bacilli. J. Exp. Med. 162:40112 Onwubalili JK, Scott GM, Smith H. 1986. Acute respiratory distress related to chemotherapy of advanced pulmonary tuberculosis: a study of two cases and review of the literature. Q. J. Med. 59:599610 Onyebujoh PC, Abdulmumini T, Robinson S, Rook GAW, Stanford JL. 1995. Immunotherapy for tuberculosis in African conditions. Resp. Med. 89:199207 Orme IM, Miller ES, Roberts AD, Furney SK, Grifn JP, et al. 1992. T lymphocytes mediating protection and cellular cytolysis during the course of Mycobacterium tuberculosis infection. Evidence for different kinetics and recognition of a wide spectrum of protein antigens. J. Immunol. 148:18996 Orme IM, Roberts AD, Grifn JP, Abrams JS. 1993. Cytokine secretion by CD4 T lymphocytes acquired in response to Mycobacterium tuberculosis infection. J. Immunol. 151:51825 Panchamoorthy G, McLean J, Modlin RL, Morita CT, Ishikawa S, et al. 1991. A predominance of the T cell receptor V gamma 2/V delta 2 subset in human mycobacteria-responsive T cells suggests germline gene encoded recognition. J. Immunol. 147:336069 Pancholi P, Mirza A, Bhardwaj N, Steinman RM. 1993. Sequestration from immune CD4+ T cells of mycobacteria growing in human macrophages. Science 260:98486 Post FA, Soule SG, Willcox PA, Levitt NS. 1994. The spectrum of endocrine dysfunction in active pulmonary tuberculosis. Clin. Endocrinol. 40:36771 Rigby WF, Yirinec B, Oldershaw RL, Fanger MW. 1987. Comparison of the effects of 1,25-dihydroxyvitamin D3 on T lymphocyte subpopulations. Eur. J. Immunol. 17:56366 Roach TIA, Chatterjee D, Blackwell JM. 1994. Induction of early-response genes KC and JE by mycobacterial lipoarabinomannans: regulation of KC expression in murine macrophages by Lsh/Ity/Bcg (candidate Nramp). Infect. Immun. 62:117684 Rook GAW. 1988. The role of vitamin D in tuberculosis. Am. Rev. Respir. Dis. 138:76870 Rook GAW, Hernandez-Pando R, Light-

75.

64.

76.

65.

77.

66.

67. 68.

78.

69.

79.

70.

80.

81.

71.

82.

72.

83.

73.

84. 85.

74.

PATHOGENESIS OF TUBERCULOSIS
man SL. 1994. Hormones, peripherally activated prohormones, and regulation of the TH1/TH2 balance. Immunol. Today 15:3013 Rook GAW, Honour J, Kon OM, Wilkinson RJ, Davidson R, Shaw RJ. 1996. Urinary steroid metabolites in tuberculosis; a new clue to pathogenesis. Q. J. Med. 88:33341 Rook GAW, Onyebujoh P, Stanford JL. 1993. TH1TH2 switch and loss of CD4 cells in chronic infections; an immuno-endocrinological hypothesis not exclusive to HIV. Immunol. Today 14:56869 Rook GAW, Steele J, Ainsworth M, Champion BR. 1986. Activation of macrophages to inhibit proliferation of Mycobacterium tuberculosis: comparison of the effects of recombinant gamma-interferon on human monocytes and murine peritoneal macrophages. Immunology 59:33338 Rook GAW, Steele J, Fraher L, Barker S, Karmali R, et al. 1986. Vitamin D3, gamma interferon, and control of proliferation of Mycobacterium tuberculosis by human monocytes. Immunology 57:15963 Rook GAW, Taverne J, Leveton C, Steele J. 1987. The role of gammainterferon, vitamin D3 metabolites and tumor necrosis factor in the pathogenesis of tuberculosis. Immunology 62:22934 Salgame P, Abrams JS, Clayberger C, Goldstein H, Convit J, et al. 1991. Differing lymphokine proles of functional subsets of human CD4 and CD8 T cell clones. Science 254:27982 Sarma GR, Chandra I, Ramachandran G, Krishnamurthy PV, Kumaraswami V, Prabhakar R. 1990. Adrenocortical function in patients with pulmonary tuberculosis. Tubercle 71:27782 Schauf V, Rom WN, Smith KA, Sampaio EP, Meyn PA, et al. 1993. Cytokine gene activation and modied responsiveness to interleukin-2 in the blood of tuberculosis patients. J. Infect. Dis. 168:1056 59 Scott GM, Murphy PG, Gemidjioglu ME. 1990. Predicting deterioration of treated tuberculosis by corticosteroid reserve and C-reactive protein. J. Infect. 21:6169 Shands JW, Sentertt VC. 1972. Endotoxin-induced hepatic damage in BCG-infected mice. Am. J. Pathol. 67:2340 Shwartzman G. 1937. Phenomenon of

283

85a.

96. 97.

86.

98.

87.

99.

88.

100.

101. 102. 103. 104.

89.

90.

91.

105.

92.

106.

93.

107.

94.

108.

95.

Local Tissue Reactivity and its Immunological, Pathological, and Clinical Signicance. New York: Hoeber. 461 pp. Siegele DA, Kolter R. 1992. Life after log. J. Bacteriol. 174:34548 Sieling PA, Chatterjee D, Porcelli SA, Prigozy TI, Mazzaccaro RJ, et al. 1995. CD1-restricted T cell recognition of microbial lipoglycan antigens. Science 269:22730 Silva CL, Lowrie DB. 1994. A single mycobacterial protein (hsp65) expressed by a transgenic antigen-presenting cell vaccinates mice against tuberculosis. Immunology 82:24448 Silver RF, Wallis RS, Ellner JJ. 1995. Mapping of T cell epitopes of the 30-kDa alpha antigen of Mycobacterium bovis strain bacillus Calmette-Guerin in puried protein derivative (PPD)-positive individuals. J. Immunol. 154:4665 74 Singhal M, Banavalikar JN, Sharma S, Saha K. 1989. Peripheral blood T lymphocyte subpopulations in patients with tuberculosis and the effect of chemotherapy. Tubercle 70:17178 Spahlinger H. 1922. Note on the treatment of tuberculosis. Lancet 1:58 Spence DP, Hotchkiss J, Williams CS, Davies PD. 1993. Tuberculosis and poverty. Br. Med. J. 307:75961 Stanford JL, Grange JM. 1994. The promise of immunotherapy for tuberculosis. Respir. Med. 88:37 Sturgill-Koszycki S, Schlesinger PH, Chakraborty P, Haddix PL, Collins HL, et al. 1994. Lack of acidication in Mycobacterium phagosomes produced by exclusion of the vesicular protonATPase. Science 263:67881 Surcel HM, Troye-Blomberg M, Paulie S, Andersson G, Moreno C, et al. 1994. Th1/Th2 proles in tuberculosis based on proliferation and cytokine response of peripheral blood lymphocytes to mycobacterial antigens. Immunology 81:17176 Swanson-Beck J, Potts RC, Kardjito T, Grange JM. 1985. T4 lymphopenia in patients with active pulmonary tuberculosis. Clin. Exp. Immunol. 60:4954 Tanaka Y, Morita CT, Tanaka Y, Nieves E, Brenner MB, Bloom BR. 1995. Natural and synthetic non-peptide antigens recognized by human gamma delta T cells. Nature 375:15558 Tobach E, Bloch H. 1956. Effect of stress by crowding prior to and follow-

284

ROOK & HERNANDEZ-PANDO

ing tuberculous infection. Am. J. Physiol. 187:399402 van-der-Pouw-Kraan TC, Boeije LC, Smeenk RJ, Wijdenes J, Aarden LA. 1995. Prostaglandin-E2 is a potent inhibitor of human interleukin 12 production. J. Exp. Med. 181:77579 Veyron P, Pamphile R, Binderup L, Touraine JL. 1993. Two novel vitamin D analogues, KH 1060 and CB 966, prolong skin allograft survival in mice. Transpl. Immunol. 1:7276 Vidal S, Tremblay ML, Govoni G, Gauthier S, Sebastiani G, et al. 1995. The Ity/Lsh/Bcg locus: Natural resistance to infection with intracellular parasites is abrogated by disruption of the Nramp1 gene. J. Exp. Med. 182:65566 Vouldoukis I, Riveros-Moreno V, Dugas B, Ouaaz F, Becherel P, et al. 1995. The killing of Leishmania major by human macrophages is mediated by nitric oxide induced after ligation of the Fc epsilon RII/CD23 surface antigen. Proc. Natl. Acad. Sci. USA 92:78048 Wilson GS, Schwabacher H, Maier I. 1940. The effect of the desensitization of tuberculous guinea-pigs. J. Pathol. Bacteriol. 50:89109 Wynn TA, Cheever AW, Jankovic D, Poindexter RW, Caspar P, et al. 1995. An IL-12-based vaccination method for preventing brosis induced by schistosome infection. Nature 376:59496 Xu S, Cooper A, Sturgill-Koszycki S, van Heyningen T, Chatterjee D, et al. 1994. Intracellular trafcking in Mycobacterium tuberculosis and Mycobacterium avium-infected macrophages. J. Immunol. 153:256878 Yong AJ, Grange JM, Tee RD, Beck JS, Bothamley GH, et al. 1989. Total and anti-mycobacterial IgE levels in serum from patients with tuberculosis and leprosy. Tubercle 70:27379 Young DB. 1992. Heat-shock proteins: immunity and autoimmunity. Curr. Opin. Immunol. 4:396400 Zack MB, Fulkerson LL, Stein E. 1973. Glucose intolerance in pulmonary tuberculosis. Am. Rev. Respir. Dis. 108:1164 69 Zuckerman SH, Shellhaas J, Butler LD. 1989. Differential regulation of lipopolysaccharide-induced interleukin 1 and tumor necrosis factor synthesis; effects of endogenous and exogenous glucocorticoids and the role of the pituitaryadrenal axis. Eur. J. Immunol. 19:301 5

109.

114.

110.

115.

111.

116. 117.

111a.

118.

112.

113.