Role of iron in experimental Mycobacterium avium infection

Abstract

Background: Acquired immunodeficiency syndrome (AIDS) patients exhibit alterations in the metabolism of iron that lead to increased deposition of this element in the tissues. Such alterations may underlie an increased susceptibility of AIDS patients to mycobacterial infections, namely by Mycobacterium avium. Objectives: The understanding of the role of iron metabolism during M. avium infections in mouse models may allow the design of new therapies based on the manipulation of iron stores. Study design: In vitro macrophage cultures and in vivo mouse studies of iron depletion and iron overload are used to assess mycobacterial multiplication and testing of the efficacy of iron depletion strategies such as the use of iron chelators. Results and conclusions: The levels of iron loading of macrophages in vitro or in vivo affect the growth of M. avium. The currently available iron chelators have poor efficacy in depleting the macrophage iron stores and, therefore, have a poor impact on the infection. Therefore, newer drugs are required that may be used in the context of in vivo infections such as in the case of affected AIDS patients.

Introduction

Mycobacterium avium is usually classified among the non-pathogenic species of the Mycobacterium genus since it only rarely infects the normal human population. However, it is one of the most frequent opportunistic infections in acquired immunodeficiency syndrome (AIDS) patients (Inderlied et al., 1993). The incidence of M. avium infection in these patients correlates strongly with the decline in the levels of circulating CD4⁺ T cells revealing a pivotal role for CD4⁺ T cells in the immune response to this pathogen, which has been largely confirmed by the data obtained from murine models of infection (Appelberg et al., 1994). The way in which CD4⁺ T cells contribute to protection against M. avium infection is not completely clear. Since M. avium infects the macrophages of the host, the macrophage-activating capacity of CD4⁺ T cells is certainly important and data from mouse models of infection suggest that IFNγ production is one, but not the only key factor (Appelberg et al., 1994, Florido et al., 1999). In humans, genetic deficiencies in the ligand-binding chain of the IFNγ receptor are associated with frequent M. avium infections (Newport et al., 1996).

Additional deficiencies found in the AIDS patients have, however, been suggested to play a role in increasing the susceptibility to M. avium. Neutropenias and neutrophil defects are frequent in HIV-infected individuals (Pitrak et al., 1993, Roilides et al., 1993, Keiser et al., 1996, Jacobson et al., 1997) and NK cell defects have been reported as well (Bonavida et al., 1986, Scott-Algara et al., 1992). Of particular relevance to the present discussion is that the AIDS patients have altered iron metabolism leading to accumulation of the metal in several tissues, including the liver (Boelaert et al., 1996). On the other hand, the mycobacteriostatic effect of apo-transferrin, present in the normal human serum is diminished in the AIDS patients’ serum (Crowle et al., 1989). These studies thus suggest an important role of iron in determining the susceptibility of the AIDS patients to M. avium. This hypothesis has been substantiated both by clinical data and by results obtained in experimental models. Two clinico-pathological studies reported a significant association between M. avium infection in AIDS patients and an increased iron load in the liver (Al-Khafaji et al., 1997) or in the bone marrow (De Monyé et al., 1999). Studies in the mouse have shown that iron overload may promote mycobacterial growth and exacerbate the M. avium infection (Dhople et al., 1996, Gomes et al., 1999a). Using an in vitro model of infection of bone marrow derived mouse macrophages, we found that adding extra iron to the culture medium increased the intra-macrophagic growth of M. avium (Fig. 1). We also tested the effect of an iron overload induced by repeated intraperitoneal injections of iron dextran in the course of an in vivo infection. When these mice were infected with M. avium, the growth of the bacterium was significantly higher than that taking place in untreated mice (Fig. 2). During the latter studies, we also found that the growth promoting effect of iron is dependent on the allele of the Nramp1 gene expressed by the mice. Natural resistance associated protein (NRAMP) 1 is a transmembrane protein expressed in endocytic vesicles of macrophages (Gruenheid et al., 1997, Searle et al., 1998). Mice bearing the mutated allele of Nramp1 (D169) are naturally susceptible to several intracellular infections including M. avium (Appelberg and Sarmento, 1990) and iron overloading of these mice leads to a modest increase in M. avium proliferation (Gomes and Appelberg, 1998). On the other hand, a similar iron overload leads to a more pronounced increase in M. avium proliferation in naturally resistant mice expressing the wild-type form of Nramp1 (G169), rendering the latter mice almost as susceptible as the former. These data have led us to speculate that the NRAMP1 protein, whose putative transporter function is otherwise unclear, is transporting iron out of the phagosome containing the M. avium bacilli (Gomes and Appelberg, 1998). Biochemical studies have supported this notion (Barton et al., 1999). Several studies have been conducted in order to search for possible mutations in the human Nramp1 gene. Although several polymorphisms have been found outside the coding regions, data on the correlation between them and susceptibility to infection are contradictory and no mutations have been found in the coding regions (Blackwell et al., 1997, Abel et al., 1998, Bellamy et al., 1998, Huang et al., 1998). Apparently, the human being bears a functional NRAMP1 protein, which could explain both its relative resistance to M. avium infection and the fact that even a mild degree of iron overload contributes to a big increase in susceptibility.

When studying the role played by iron in M. avium infections, we have to consider that iron can influence both the host macrophage and the intracellular pathogen. Macrophages are phagocytic, long-lived cells that are endowed with a multitude of antimicrobial mechanisms, which can be further activated by exposure to cytokines, namely IFNγ and TNFα. Several studies suggest that the addition of iron decreases or abolishes the antimicrobial activity induced by IFNγ on macrophages (Byrd and Horwitz, 1989, James and Glaven, 1989, Lane et al., 1991, Mauel et al., 1991, Gebran et al., 1994), although no such data exist for M. avium. Furthermore, iron was shown to inhibit the activation of iNOS, presumably one of the most important mediators of intra-macrophagic microbial killing (Weiss et al., 1994), but the growth of M. avium inside macrophages is not inhibited by NO (Appelberg and Orme, 1993, Doherty and Sher, 1997, Gomes et al., 1999b). One single instance in which iron is said to increase the macrophages’ antimicrobial capacity is the generation of reactive oxygen species. Macrophages possess a membrane oxidase capable of producing high amounts of superoxide in response to different stimuli (Rotrosen, 1992). Iron present inside phagosomes could presumably catalyze the transformation of this radical into more toxic species. However, M. avium is resistant to reactive oxygen species as evidenced by the fact that oxidase knock-out animals are not more susceptible to this infection (Segal et al., 1999). Also, all the data presented above indicate that the increase in the availability of iron leads to an increase in M. avium growth, contrary to what would be expected if iron was involved in M. avium killing. We, thus, favor the hypothesis that iron promotes the growth of M. avium directly, when made available in the phagosome, rather than interfering with the macrophages’ antimicrobial capacity. In the context of the host defense against M. avium infection, any mechanism that allows the macrophage to withdraw iron from the Mycobacterium-containing phagosome will thus be beneficial. The NRAMP1 protein may be one of such mechanisms, but others may exist that remain to be uncovered, namely alterations induced by IFNγ. It is known that IFNγ treatment causes a decrease in the expression of transferrin receptors at the macrophage surface (Byrd and Horwitz, 1989). This could cause a decrease in the amount of iron available for intracellular M. avium, according to data obtained with M. tuberculosis, which indicates that exogenously added transferrin is accessible to mycobacteria-containing phagosomes (Clemens and Horwitz, 1996). However, the expression of transferrin receptors by macrophages is low (Testa et al., 1991) and their iron acquisition from low molecular weight chelates is much more efficient than from transferrin (Olakanmi et al., 1994). It can also be predicted that in vivo the main contribution for intra-macrophagic iron is probably erythrophagocytosis (Gordeuk et al., 1994). Another alteration induced by IFNγ in macrophages is the increase in ferritin synthesis (Weiss et al., 1993, Recalcati et al., 1998), which can also contribute to a decrease in available iron. The question remains as to how iron normally reaches the bacterial phagosome. We can speculate that the Mycobacterium-produced siderophores may play a role in this process, since they have been reported to be capable of extracting iron from transferrin and ferritin (Gobin and Horwitz, 1996).

If iron excess plays a role in the increased susceptibility to M. avium in humans, then iron chelating therapy should be considered as part of the anti-mycobacterial therapy. In vitro testing of several iron chelating compounds has shown that desferrioxamine and hydroxybenzylethylenediamine-diacetic acid (HBED) are moderately effective in reducing M. avium growth in cultured macrophages (Gomes et al., 1999a and Fig. 1). However, when testing these compounds in vivo, in M. avium infected mice, we found that they have a very limited effect (Gomes et al., 1999a). Therefore, to test the concept that the restriction of iron would be beneficial to the control of M. avium infections, we have tested the effects of a low-iron diet. C57B1/6 mice were kept in a low iron diet (6.7 mg/kg), starting 2 weeks before the infection and during the complete duration of the infection. The results in terms of bacterial growth are shown in Fig. 2. The results suggest that it is possible to affect the proliferation of M. avium by restricting its access to iron, since a clear inhibition of the M. avium growth is seen in the organs of mice rendered iron deficient. The lack of effect of currently available chelators is probably the result of a lack of iron depletion efficiency, which is indicated by the fact that, contrary to mice fed with low iron diet, DFO or HBED-treated mice showed no decrease in their hematocrit, transferrin saturation, or total serum iron (Gomes et al., 1999a). To assess the role of iron chelators in the context of opportunistic infections by M. avium in AIDS patients, more adequate models should probably be used, such as iron-overloaded animals with T cell deficiencies. Also, new molecules are required, namely those specifically designed to be taken up by macrophages and to deplete iron from the intracellular compartments where mycobacteria are proliferating, the phagosomes.

In summary, our data give support to the notion that manipulating the status of iron metabolism during opportunistic infections in the AIDS patients may be of prophylactic and/or therapeutical value and that the search for new, more effective iron chelating compounds is warranted.