Cell-mediated, but not antibody-mediated, immune responses protect humans against certain pathogens that produce chronic diseases such as leishmaniasis. Effective vaccination against such pathogens must therefore produce an immunological "imprint" so that stable, cell-mediated immunity is induced in all individuals after natural infection. BALB/c mice "innately susceptible" to Leishmania major produce antibodies after substantial infection. In the present study, "susceptible" mice injected with a small number of parasites mounted a cell-mediated response and acquired resistance to a larger, normally pathogenic, challenge. This vaccination strategy may be applicable in diseases in which protection is dependent on cell-mediated immunity.

Leishmania major causes cutaneous leishmaniasis in mice and man. Infection of mice with relatively low or high numbers of parasites leads respectively to parasite containment, associated with a Th1, cell-mediated response, or progressive disease, associated with a Th2, antibody response in all circumstances studied. These include different parasite strains, different routes of infection, and different hosts previously classified as susceptible, resistant or of intermediate susceptibility. This dose dependency appears to reflect a general rule. We argue that this rule may allow the design of a vaccination strategy that is effective among a genetically diverse population, and that it imposes severe constraints upon proposals for the nature of the "decision criterion" determining whether antigen induces a Th1 or Th2 response.

BALB/c mice are susceptible to a high-dose infection of the protozoan Leishmania major, which induces a parasite-specific antibody, Th2-like response, exclusive of a significant and protective cell-mediated Th1 component. We have shown, in contrast, that infection with a low number of parasites induces cell-mediated immunity exclusive of antibody production, and results in resistance to substantial subsequent high-dose infection. Low-dose exposure thus constitutes effective vaccination. In the present study, we analyze lymphokine production by parasite-specific T cells from those low-dose exposed, resistant mice and from normal, susceptible mice following high-dose infection. Two findings stand out. First, the parasite-specific T cells in mice rendered resistant appear not to be in an activated, effector state at the time of parasite challenge, as assessed by lack of lymphokine production on short-term stimulation with parasite antigens, but to be rather in a memory state. Second, the ratio of parasite antigen-dependent production of interferon-gamma to that of interleukin-4 by spleen cells of low-dose exposed and normal mice upon high-dose challenge takes a dramatically different course. This ratio is similar in both groups of mice shortly after challenge, but increases dramatically in the resistant and declines dramatically in the control mice over a period of weeks, such that these ratios differ by about 60-fold 12 weeks after the high-dose challenge. In addition, we show that a similar state of resistance occurs following low-dose infection with a more virulent strain of L. major. In toto, our observations suggest that resistance may be generally achievable by low-dose exposure and may be associated with a memory state which, when activated by parasite challenge, results in the evolution of the response over weeks such that the protective, Th1 component becomes ever more dominant over the Th2 component.

Significant levels of IgG3 and IgG4 and high levels of IgG1 leishmania-specific antibody differentiated the immune states in 10 patients with visceral leishmaniasis from those of virtually all 20 drug-cured and 18 subclinically infected subjects, whereas the level of IgG2 antibody was nondiscriminating. The most extreme "subclinically infected" outlier subsequently developed disease. Overall, the immune states in subclinically infected and drug-cured persons were mutually indistinguishable but were readily distinguished from those of patients. These findings may have implications for the immunologic mechanism underlying drug cure in visceral leishmaniasis.

Most people infected by Mycobacterium tuberculosis, about 90%, contain the pathogen and are healthy. Most investigators have concluded that pathogen-specific Th1 cells contribute to protection. Pulmonary tuberculosis, the most prevalent form of disease, is associated with destructive granulomas, the formation of which also appears to involve Th1 cells. In what sense then do the two Th1 components of the response, in healthy infected individuals and patients, differ? An insight into this question might provide clues for attaining effective vaccination and better treatment. We approached this question by examining the relative prevalence of different IgG isotypes among anti-mycobacterium-specific antibodies in patients and healthy infected individuals as a surrogate marker for the Th1/Th2 phenotype of the response. Our observations lead us to agree that healthy infected individuals generate a predominant Th1 response. Our observations also lead us to propose that many patients make a similar kind of response as healthy infected individuals, but that this response is too weak to contain the infection. We refer to such individuals as having type I tuberculosis. Other patients appear to have a greater and detrimental Th2 component to their immune response than that of healthy infected individuals. We refer to these individuals as having type II tuberculosis. This proposal that there are two types of tuberculosis, reflecting two distinct types of failure by the immune system, will, if correct, be pertinent to vaccine design, treatment of tuberculosis and in making further progress in our understanding the genetics of susceptibility to M. tuberculosis.