The galvanizing of Mycobacterium tuberculosis: An antimicrobial mechanism

Summary

Evolving under constant threat from invading microbes, macrophages have acquired multiple means of killing bacteria. In this issue of Cell Host & Microbe, Botella and colleagues describe a novel anti-microbial mechanism based on elevated levels of intraphagosomal Zn²⁺ and the corresponding induction of bacterial genes to ameliorate this host-derived stress.

There are few inter-organismal interactions that are more intimate than the one between a pathogen that causes a chronic infection and its host. Such an interaction molds the evolution of both partners so that all anti-microbial strategies acquired by the host are countered by avoidance or neutralization mechanisms developed by the pathogen. The “success” of either organism depends on the relative balance of these responses.

The paper by Botella and colleagues (2011) documents an anti-microbial mechanism that is induced in the macrophages upon infection by Mycobacterium tuberculosis (Mtb). The macrophage releases free Zn²⁺ from intracellular stores, which is then pumped into the newly-formed phagosome raising the intra-phagosomal Zn²⁺ concentration to levels that would limit growth of Mtb. In response, Mtb up-regulates expression of the P-type ATPase-encoding gene ctpC, which regulates the intra-bacterial levels of Zn²⁺ through efflux of the metal ion. Mtb deficient in expression of this transporter exhibited impaired intracellular survival. The key steps in this process are summarized in Figure 1.

Figure 1.

This diagram illustrates the sequence of events that lead to increased Zn²⁺ concentration within the Mycobacterium tuberculosis-containing phagosome, and the subsequent up-regulation of both host and bacterial genes in response to free Zn²⁺. During phagocytosis of Mtb the NADPH Oxidase complex is formed on the phagosomal membrane and generates reactive oxygen intermediates (ROIs). These ROIs signal back on the host cell and induce release of free Zn²⁺ from intracellular stores, such as the mitochondria. The free Zn²⁺ is transported into the Mtb vacuole and other endosomal compartments through unknown Zn²⁺ transporters. Within the phagosome, this increase in free Zn²⁺, which is potentially toxic to Mtb, is countered through the up-regulation of the bacterial Zn²⁺ efflux pump CtpP. Within the host cell, the increase in cytoplasmic Zn²⁺ induces the migration of metal regulatory transcriptional factor 1 (MTH1) to the nucleus where it up-regulates transcription of genes encoding metallothioneins and the Zn²⁺ efflux pump ZnT1. ZnT1 is assembled in the plasmalemma and pumps free Zn²⁺ out of the infected cell.

Using Fluozin-3 (FZ3), a probe specific for free Zn²⁺, the authors demonstrated that Zn²⁺ is rapidly released from intracellular stores following infection with Mtb. A similar mobilization of intracellular Zn²⁺ stores was also observed upon infection with E. coli, suggesting that the macrophage was being stimulated through its innate immune receptors, such as Toll-like receptors. Intriguingly, known inhibitors of the NADPH oxidase, which generates the reactive oxygen intermediates (ROIs) induced during the phagocytosis of particles such as Mtb, blocked the release of free Zn²⁺, implicating reactive oxygen intermediates in the release of free Zn²⁺ (Andrews, 2000). The exact pathways leading to the mobilization of these intracellular stores of Zn²⁺ would be an interesting area for future study. Recently, it has been suggested that Mtb actively down-regulates the production of ROIs through bacterial NADH dehydrogenase activity, which has the potential to both reduce reactive oxygen species and the release of Zn²⁺ from the host cell mitochondria (Miller et al., 2010).

The authors exploited some elegant histochemistry to localize those regions enriched for free Zn²⁺ in both the macrophages and the intracellular bacteria at the ultrastructural level. Stoltenberg and colleagues developed an autometallographic technique whereby fixation in the presence of sodium sulfide induces the formation of zinc-sulfur crystals that can be enhanced by silver (Stoltenberg et al., 2005). Using this method Botella and co-workers observed that, following infection, silver crystals were visible in the host cell within the mitochondria, within endocytic compartments and associated with the inner face of the plasmalemma (Botella et al. 2011). In the bacterium they identified three distinct patterns of localization. Silver crystals were seen associated with the surface of the bacterium, within a region that likely corresponds to the periplasmic space, and within the cytoplasm of the bacterium. This latter localization was seen more frequently in bacterial cells that appeared compromised, implying the increased Zn²⁺ concentration correlated with bacterial death. This hypothesis was substantiated further by comparison between wild-type Mtb and a mutant bacterial strain deficient in expression of CtpC, a P-ATPase transporter associated with the efflux of Zn²⁺. When macrophages were infected with both strains there were significantly more zinc-sulfur crystals seen in the cytosol of the mutant.

The increased concentration of free Zn²⁺ in the bacteria-containing phagosomes is consistent with a previous study by Wagner and colleagues who documented elevated levels of metal ions in Mtb- and M. avium (Mac)-containing phagosomes as measured by hard x-ray microprobe (Wagner et al., 2005). The levels of both Fe²⁺ and Zn²⁺ exhibited interesting deviations from the extracellular concentrations. The Mtb and Mac-containing phagosomes exhibited elevated levels of Fe²⁺. This is consistent with previous reports demonstrating the cycling of the iron-binding protein transferrin through the Mtb-containing vacuole (Clemens and Horwitz, 1996; Sturgill-Koszycki et al., 1996), and transcriptional profile studies that indicate that Mtb is in an iron-replete environment (Homolka et al., 2010; Tailleux et al., 2008). Mtb defective in the synthesis of iron-chelating siderophores do not exhibit this increase suggesting that the bacterium is responsible for sequestering Fe²⁺ within the phagosome (Wagner et al., 2005). Treatment of macrophages infected with Mac with TNF-α, which impairs bacterial survival, neutralized the increase in intra-vacuolar Fe²⁺. In contrast, the concentration of Zn²⁺ was elevated most markedly in the phagosomes containing the avirulent mutant Mtb defective in siderophore production and in Mac-containing phagosomes following activation of the host cells with TNF-α (Wagner et al., 2005)33 From these data it is clear that Fe²⁺ concentration correlates positively with bacterial fitness, while Zn²⁺ concentration correlates negatively with bacterial fitness. However, the significance of these observations was not apparent at the time of the earlier publication by Wagner and colleagues but is brought sharply into focus by the current study of Botella and coworkers, who identify Zn²⁺-loading as a host-mediated anti-microbial response. The route by which Zn²⁺ gets into the phagosome remains to be determined.

Previous transcriptional profiling of both Mtb and the host cell in macrophage infection models had demonstrated the strong and coordinated up-regulation of expression of genes linked to heavy metal poisoning or the response to excess free metals in both cells (Homolka et al., 2010; Tailleux et al., 2008). Upon entry into macrophages, Mtb markedly up-regulated expression of several P-type ATPases commonly associated with the efflux of free metal ions such as Zn²⁺ or Cu²⁺, and the induction of expression of two of these genes, cptG and cptC, was enhanced by increased Zn²⁺ concentration in vitro. Mutant Mtb deficient in the expression of one of these putative transporters, CtpC, accumulated Zn²⁺ in culture, and as discussed previously, inside macrophages. On the host cell side, the macrophage showed enhanced expression of metal toxicity-linked genes such as the metallothioneins, and the Zn²⁺ exporter ZnT1. In addition, the infected cells demonstrated translocation of the metal-regulatory transcription factor 1 (MTF1) from the cytosol to the nucleus where it binds to the regulatory regions of known genes involved in controlling metal-mediated stress (Andrews, 2001). These data argue strongly that the release of free Zn²⁺, induced upon bacterial infection, is impacting the biology of both organisms.

All these data generate a compelling story of a novel anti-microbial mechanism and Mtb’s response to ensure its continued survival. However, one area where this story is less clear is the phenotype of the ctpC-deficient mutant. This mutant exhibited a reduced capacity to survive in human macrophages in culture and the phenotype could be complemented upon ectopic expression of ctpC. In contrast, the mutant lacked any phenotype upon infection of wild-type Balb/C mice. This indicated that either the bacteria had compensatory mechanisms that were not invoked in in vitro macrophage infections, or that there was a host difference between the mouse in vivo infections and the human macrophage infections in culture. There is a history of literature documenting the disparities between the induction of anti-microbial circuits in human versus mouse macrophages and this may simply be another divergence of note.

To demonstrate that the induction of free Zn²⁺ and its delivery to the phagosome has broad significance as an anti-microbial response, the authors looked at the survival profiles of wild-type E. coli versus a mutant strain in which the zinc efflux ATPase, ZntA, had been inactivated. Upon uptake into host phagocytes, both strains induced release of free Zn²⁺, which was seen to concentrate in the bacteria-containing vacuoles. The rate of killing of the wild-type bacteria was markedly slower than the zntA-deficient mutant. These data provide an independent validation of the hypothesis that the generation of free Zn²⁺ upon stimulation by microbial insult, coupled with its delivery into the phagosome, represents a novel strategy to eradicate intracellular pathogens.

Several obvious questions remain. How do the ROI’s induce release of Zn²⁺? Which transporters deliver the Zn²⁺ into the phagosome? How does Zn²⁺ intoxicate these bacteria? And what is the basis of the disparity between the human macrophage infections and the murine infection? However, none of these questions detract from the appeal of these data and the underlying hypothesis, whereby the concentration of the essential metal ion, Zn²⁺, is manipulated by the host to determine the success or failure of an infection.