Expression of NADPH Oxidase-Dependent Resistance to Listeriosis in Mice Occurs during the First 6 to 12 Hours of Liver Infection

Abstract

Wild-type mice inoculated with Listeria monocytogenes intravenously were capable of reducing the bacterial load in their livers by 90% within 6 h. In contrast, mice with deletions of the gene for NADPH oxidase were incapable of expressing this early oxygen-dependent anti-Listeria defense and consequently showed higher levels of liver infection at later times.

The oxygen-dependent antimicrobial mechanism of macrophages and neutrophils relies on activation of NADPH oxidase that catalyzes the transfer of electrons from NADPH to molecular oxygen, resulting in the formation of superoxide anion and a number of downstream toxic oxidants (1). Resistance to infection with the facultative intracellular bacterium Listeria monocytogenes in mice depends on the antimicrobial mechanisms of both neutrophils and macrophages (7, 8, 10). Neutrophils are believed to express their anti-L. monocytogenes function early in infection, whereas macrophages express this function mainly at a later time after being activated by L. monocytogenes-specific T cells at sites of infection. Evidence that neutrophils are essential participants in antilisteriosis resistance is based on the observation (3, 8) that L. monocytogenes infection is severely exacerbated in mice that have been depleted of neutrophils by treatment with a granulocyte-specific monoclonal antibody. According to one study with this antibody (3), the infection-exacerbating action of neutrophil depletion is evident in the liver as early as 12 h of infection, whereas according to another study (8), neutrophils express their function at 48 h. In either case, the antibacterial function of neutrophils would be expected to depend to a significant extent on their ability to generate reactive oxygen and reactive oxygen intermediates. It has been shown (5), in support of this expectation, that mice with deletions of the gene for the gp91^phox component of NADPH oxidase (phox^−/− mice) allow significantly more L. monocytogenes growth in their livers and spleens than wild-type (WT) mice, as evidenced by a higher level of infection at 48 h postinoculation. However, an increased level of listeriosis in phox^−/− mice at 48 h of infection need not by itself represent evidence for a loss of neutrophil function, given that macrophages also depend on NADPH oxidase to destroy L. monocytogenes (9). Moreover, a higher level of listeriosis in phox^−/− mice at 48 h of infection could be a consequence of a higher level of infection at a much earlier time. It was deemed important to determine, therefore, whether mice incapable of making NADPH oxidase are deficient in the ability to control L. monocytogenes infection during the first few hours of infection, shortly after the pathogen is removed from the circulation by macrophages of the liver and spleen.

This was investigated by monitoring the growth of L. monocytogenes in the livers and spleens of C57BL/6 WT mice and C57BL/6 mice with deletions of the gene for gp91^phox (from breeding stock supplied by Carl Nathan, Cornell University Medical School, New York, N.Y.) inoculated intravenously (i.v.) with 2 × 10³, 8 × 10³, or 8 × 10⁴ CFU of L. monocytogenes strain EGD (originally obtained from George Mackaness, Trudeau Institute, Saranac Lake, N.Y.). L. monocytogenes was cultured and prepared for i.v. inoculation as previously described (3). At progressive times after inoculation five mutants and five WT mice were sacrificed, their livers and spleens were harvested, and each organ was homogenized in cold phosphate-buffered saline. Tenfold serial dilutions of the homogenates were plated on nutrient agar, and colonies were enumerated after 24 h, as previously described (3).

The course of infection in the livers and spleens of WT and phox^−/− mice given a sublethal i.v. inoculum (2 × 10³ CFU) of the pathogen is shown in Fig. 1 where it can be seen that in both types of mice, most (95%) of the inoculum was taken up by the liver, as evidenced by the number of CFU in this organ 30 min postinoculation. Most of the remainder of the inoculum was taken up by the spleen. In WT mice, the number of L. monocytogenes CFU in the liver declined by approximately 1 log during the first 6 h of infection, after which the organism multiplied log linearly for the following 18 h. After this time, bacterial growth was controlled and L. monocytogenes numbers leveled off until 48 h, when the experiment was terminated. In the livers of phox^−/− mice, by contrast, there was no significant reduction in L. monocytogenes number during the first 6 h. After 6 h the shape of the bacterial growth curves in the livers of WT and phox^−/− mice was the same, except that bacterial numbers in the latter mice were 1 log higher. The same basic result was obtained when WT and phox^−/− mice were given higher lethal inocula of L. monocytogenes (Fig. 2), although in this case the earliest time at which bacterial growth was monitored was 12 h. In the spleen, there was no reduction in bacterial number in WT or phox^−/− mice (Fig. 1 and 2) during the first 6 to 12 h of infection. However, L. monocytogenes grew slightly faster in the spleens of phox^−/− mice between 12 and 24 h postinoculation, resulting in a higher level of infection at 24 and 48 h.

FIG. 1.

Growth of an inoculum of 2 × 10³ CFU of L. monocytogenes, given i.v., in the livers and spleens of WT and phox^−/− mice over 48 h of infection. The bacterial load in the liver was reduced by approximately 1 log (P = 0.001, according to Student's t test) during the first 6 h in the livers of WT mice but not in the livers of phox^−/− mice. There was no immediate reduction in bacterial number in the spleens of mice of either type. Values are means ± standard deviations for five mice.

FIG. 2.

Growth of inocula of 8 × 10³ and 8 × 10⁴ CFU of L. monocytogenes in the livers and spleens of WT and phox^−/− mice, showing an increased level of infection in the liver, but not in the spleen, at 12 h of infection. Values are means ± standard deviations for five mice.

These results show that the higher level of infection in the livers of phox^−/− mice at 24 and 48 h postinoculation was a consequence of failure to destroy most of the L. monocytogenes inoculum during the first 6 h. Given the histological demonstration (2) that the only cells in the liver in which L. monocytogenes can be shown to reside 30 min after administration of a 5 × 10⁸ inoculum of the organism i.v. are liver macrophages (Kupffer cells), it seems likely that these cells were mainly responsible for reducing the bacterial load. It has been shown (9) in support of this suggestion that peritoneal macrophages from WT but not from phox^−/− mice are capable of rapidly destroying a L. monocytogenes challenge infection in vitro. It seems reasonable to assume that liver macrophages kill this pathogen by the same means. The possibility cannot be excluded, however, that neutrophils may have contributed to the early NADPH oxidase-dependent reduction of liver infection by killing L. monocytogenes organisms that were adherent to the plasma membranes of liver macrophages, as proposed by Gregory and Wing (6). However, the present study shows that L. monocytogenes grows progressively in the livers of control and phox^−/− mice, between 6 and 24 h of infection, when massive neutrophil accumulation is known to occur at sites of infection (2, 3).

The findings presented here do not support the proposal by others (5) that exacerbation of listeriosis in the livers of phox^−/− mice is the result of a deficiency in the bactericidal function of neutrophils. According to one published study (3), mice depleted of neutrophils show a 1-log-higher level of infection in the liver, but not in the spleen, as early as 12 h postinoculation. Since this is similar to the results obtained here with phox^−/− mice, it could point to the early destruction of L. monocytogenes in the livers of WT mice by the oxygen-dependent antimicrobial function of neutrophils. However, the increased level of liver infection in neutrophil-depleted mice was interpreted not as showing that neutrophils kill L. monocytogenes during the first 12 h of infection but as showing, on the basis of histological evidence, that neutrophils function in antilisteriosis defense by lysing infected liver parenchymal cells after 12 h, thereby releasing the pathogen from L. monocytogenes growth-permissive cells for ingestion by phagocytic cells, which are not growth permissive. However, this interpretation was based on a histological study of livers harvested at 24 h of infection, rather than on livers harvested before 12 h. It remains to be shown that the lytic function of neutrophils is expressed against L. monocytogenes-infected hepatocytes during the first 6 h of liver infection. Moreover, because it is shown here that the growth rate of L. monocytogenes in phox^−/− mice was very similar to its growth rate in WT mice after 6 h of infection, it seems unlikely that NADPH-dependent neutrophil- or macrophage-mediated defense was expressed from 6 h on. It is worth considering the possibility, therefore, that the infection-exacerbating action of antibody-mediated neutrophil depletion, as seen at 12 h of infection, might be based on temporary loss of liver macrophage function caused by ingestion of large numbers of dead neutrophils. This would result in higher levels of infection at later times. This explanation would also apply to the listeriosis-exacerbating action of antineutrophil antibody given at 24 h of infection (4), because liver lesions contain large very numbers of neutrophils at that time (3).

The proposal (5) that oxygen-dependent neutrophil listericidal function is not expressed in the liver until 48 h of infection is also not supported by the results of this study. According to the results presented here, the higher level of liver infection in phox^−/− mice at 48 h was a consequence of a higher level of infection at 6 h. The role of NADPH oxidase-dependent defense after 48 h of infection is under study in this laboratory.

The reason for the absence of the early mechanism of defense in the spleen is not known. Presumably the splenic macrophages that remove L. monocytogenes from the circulation are devoid of the capacity to kill this pathogen. Exactly where these macrophages reside in the spleen is not known, although preliminary results in this laboratory (unpublished) suggest that L. monocytogenes is removed mostly by marginal-zone macrophages that are known to be functionally different from other macrophages (11). The absence of an early anti-L. monocytogenes defense in the spleen in neutrophil-depleted mice has also been noted (3). It is possible that the increased rate of growth of L. monocytogenes in the spleens of phox^−/− mice observed between 12 and 24 h of infection in the study presented here was due to the absence of NADPH oxidase-dependent function of neutrophils, macrophages, or both.