Listeriolysin O (LLO) is a secreted pore-forming toxin of the facultative intracellular bacterium Listeria monocytogenes. We assessed the ability of a murine anti-LLO mAb to affect the course of infection in mice challenged with Listeria. This mAb was previously shown to be capable of neutralizing LLO-mediated pore formation in vitro, and here we show that the passive administration of this Ab to mice before infection provides increased resistance. Mice treated with the mAb were protected from a lethal challenge with virulent Listeria and showed a significant reduction in Listeria burden during the first hours to days postinfection. These effects of the Ab were independent of host B or T cells, since treatment with the mAb provided enhanced resistance to SCID mice. The titer of anti-LLO Abs during the regular infection of mice with Listeria was found to be low to negative.
The mouse model of Listeria infection has long been used to study the mechanisms of host defense against intracellular bacteria, a process known to require components of both the innate and adaptive immune responses (1). Early studies by G. B. Mackaness and others indicated that serum taken from mice following sublethal infection provided no resistance against Listeria when passively transferred to naive mice (2). At the time of these studies, the specificity of anti-Listeria Abs in this polyclonal serum was not investigated. Because some investigators had also noted the failure of antisera to influence resistance against tuberculosis, the attention was then focused on cellular mechanisms of resistance (reviewed in Ref. 3). Indeed, studies in both experimental models led to the concept of cell-mediated immunity, with the requirement for T lymphocytes and activated macrophages in the acquired resistance to these intracellular bacteria (2, 4, 5, 6).
In the experiments described here, we demonstrate that an Ab to listeriolysin O (LLO)3 provides resistance to Listeria infection. LLO, the Ag recognized by the A4-8 mAb, is a 58-kDa secreted protein of Listeria monocytogenes and an important virulence factor (reviewed in Ref. 7). LLO is a member of the family of thiol-activated pore-forming toxins secreted by numerous Gram-positive bacterial species (8). LLO mediates escape of Listeria from the phagosome of an infected cell, allowing the organism to access the cytosol, where it replicates rapidly and avoids the host cell’s microbicidal actions (9, 10). LLO has also been shown to have numerous exotoxic actions when applied to cells in vitro, although the significance of these cellular responses during in vivo pathogenesis has not been established (11, 12, 13, 14, 15, 16, 17, 18).
Materials and Methods
CB.17 and CB.17-SCID mice were maintained and bred under SPF conditions in the Washington University mouse facility. For infection with Listeria, mice of either sex were used, between 8 and 16 wk of age.
All experiments utilized L. monocytogenes strain EGD. The LD50 of this strain in CB.17 or CB.17-SCID mice is ∼1 × 103 organisms. Bacteria was stored as frozen glycerol stocks and thawed once before dilution into pyrogen-free saline for injection into mice.
mAbs and in vivo use
mAbs utilized in this study were the A4-8 mAb (murine anti-LLO, IgG1) and the irrelevant control GIR.208 mAb (murine anti-human IFN-γ receptor, IgG1, a gift of Dr. Robert Schreiber, Washington University, St. Louis, MO). A4-8 was generated by the immunization of mice with purified LLO plus Freund’s adjuvant (19). This mAb was chosen because it had been shown to bind LLO in solution and block both membrane binding by the toxin and subsequent lysis of RBC in vitro. mAbs were purified from ascites using standard methods on protein A-Sepharose (Sigma, St. Louis, MO) and were shown to contain less than 10 endotoxin units per milligram protein by the QCL-1000 endotoxin quantitation kit (BioWhittaker, Walkersville, MD). Ab concentration was determined by the measurement of absorbance at OD 280 (1%:13.5). In most experiments, mice were injected with 1 mg mAb per mouse i.p., 1 day before i.p. infection with Listeria. Following infection, mice were followed for survival for 14 days or sacrificed at various time points postinfection to determine organ Listeria burden. This was done by aseptic removal of the spleen and/or liver, followed by the homogenization of each organ in PBS plus 0.05% TX-100. Serial dilutions of homogenate were plated on brain heart infusion agar, and bacterial CFU were assessed after overnight growth at 37°C.
Assessment of Ab responses to Listeria infection
CB.17 mice were infected three times (intervals of 2 to 4 wk before reinfection) with Listeria, at doses of 5 × 102, 1 × 103, and 1 × 104 organisms i.p. Serum Ab responses to whole Listeria Ag and LLO were measured using the following ELISA protocols. To test for Ab to whole Listeria Ag, Nunc Maxisorp plates (Nunc, Roskilde, Denmark) were coated overnight at 4°C with 1 × 107/well live Listeria in 0.1 M carbonate buffer (pH 8.5). Plates were washed at each step with PBS plus 0.05% Tween 20, blocked with PBS plus 1% BSA for 1 h at room temperature, incubated with serum samples diluted in PBS plus 1% BSA for 2 h at room temperature, and then incubated with secondary Ab (goat anti-mouse IgG-peroxidase; Boehringer-Mannheim, Indianapolis, IN) diluted in PBS plus 1% BSA for 1 h at room temperature. Plates were then treated with 1 mM ABTS in citrate buffer with 0.05% H2O2, and OD 415 was measured. To test for Ab to LLO, an identical ELISA was performed, this time coating the plate initially with 100 ng/well histidine-tagged recombinant LLO (produced in Escherichia coli, a gift of Dr. Daniel Portnoy, University of California, Berkeley, CA (M. M. Gedde and D. A. Portnoy, unpublished observations)). To test whether polyclonal anti-Listeria Abs could provide protection against Listeria infection in vivo, serum was obtained from control CB.17 mice and Listeria immune mice (the same mice whose serum was tested above, serum taken 2 wk following a fourth infection with 1 × 104 Listeria i.p.). Serum was pooled, and crude Ig was obtained by precipitation in 50% saturated ammonium sulfate using standard methods. Naive CB.17 mice were administered 2 mg/mouse of control Ig or immune Ig 1 day before infection with 1 × 104 Listeria i.p. Mice were sacrificed at day 2 postinfection, and organ Listeria burden was assessed as described above.
Data were analyzed using GraphPad PRISM software (Version 2.0, GraphPad Software, San Diego, CA). Survival curves were compared using the logrank test, and CFU determinations were compared using the Mann-Whitney test.
Several experiments showed that the anti-LLO mAb A4-8 provided resistance to Listeria, as noted by both the number of organisms in the spleen and liver, and by survival of the mice after infection with relatively high infectious doses. The amount of A4-8 that affected protection was 1 mg per mouse (weight of mice ∼25 g). Lower amounts resulted in much less of an effect. In no instance did we find an effect of the control Ab of the same isotype. Representative experiments are indicated in the figures. Fig. 1 A shows the survival of CB.17 mice administered either the anti-LLO mAb A4-8 or an irrelevant isotype control mAb 1 day before infection with a relatively high dose, 7.5 × 103 Listeria i.p. A second control group was also included, which received no treatment before infection with Listeria. Mice were followed for 14 days postinfection, with the majority of control mice succumbing to infection, mostly on days 4 through 7 (GIR.208 group: 9/10 dead; untreated group: 10/12 dead). The majority of A4-8-treated mice survived infection (3/10 dead), indicating a protective effect of the anti-LLO mAb. In a separate experiment, CB.17 mice were treated with A4-8 or control mAb 1 day before infection with 1 × 103 Listeria i.p. Survivors of this primary infection (A4-8 group: 2/10 dead; GIR.208 group: 5/9 dead) were then challenged with a normally lethal dose of Listeria (5 × 104), and all but one mouse (originally GIR.208 treated) survived this secondary infection, indicating that A4-8 treatment protected and did not affect the ability of mice to generate a memory response to Listeria during primary infection.
To determine the requirement for B or T cells during Ab-mediated resistance in vivo, a similar experiment was performed in SCID mice, devoid of these cell types due to a genetic mutation in the DNA-PK gene. CB.17-SCID mice were administered A4-8 or the control mAb 1 day before infection with 2 × 103Listeria i.p., and then readministered the same mAb on days 5 and 10 postinfection. At this dose of Listeria, approximately half of the control mice survived infection through day 14 (6/10 dead), while again the majority of A4-8-treated mice survived to this time point (1/9 dead) (Fig. 1 B). Since SCID mice are known to develop a chronic infection with Listeria due to the lack of T cells, we assessed the surviving mice in this experiment for the presence of Listeria in the spleen and liver on day 14 (20). Both control and anti-LLO-treated survivors were found to harbor similar levels of Listeria in these organs (geometric mean CFU of controls: 9.0 × 102/spleen, 1.3 × 103/liver; of A4-8 treated: 3.3 × 103/spleen, 2.3 × 103/liver), indicating that, while the A4-8 mAb provided enhanced survival in Listeria-infected SCID mice, it was not sufficient in the absence of T cells to bring about sterilizing immunity.
Anti-LLO provided resistance through the limitation of organ Listeria growth during infection. CB.17 mice were administered A4-8 or the control mAb 1 day before infection with 1 × 104 Listeria i.p. At day 2 postinfection, the mice treated with anti-LLO mAb had ∼100-fold fewer CFU in the spleens and livers, a very dramatic reduction and the likely explanation for how the Ab provided enhanced survival in the previous experiments (Fig. 2, B and C). When a lower dose of Listeria was used (1 × 103) and mice were examined at day 4 postinfection, A4-8-treated mice had no detectable bacteria in the spleen or liver (limit of detection 100 organisms/organ), while control mAb-treated mice had geometric mean CFU of ∼104 Listeria/organ. A4-8 also affected Listeria CFU at day 2 postinfection in mice infected with Listeria i.v. (data not shown). In preliminary studies, a nonneutralizing anti-LLO mAb was unable to affect Listeria CFU during Listeria infection, suggesting that neutralization of LLO is important for Ab-mediated resistance. At day 2 postinfection following a dose of 1 × 104 Listeria i.p., mice treated with the E4-3 nonneutralizing mAb to LLO (see Ref. 19) had geometric mean CFU of 1.0 × 105Listeria/organ, while the control mAb-treated mice had geometric mean CFU of 2.8 × 105 Listeria/organ. In this same experiment, mice treated with A4-8 again had ∼100-fold fewer CFU/organ, having a geometric mean CFU of 2.5 × 103 Listeria/organ.
The effects of the A4-8 mAb could be detected at 6 h postinfection. In pilot experiments, it was determined that to recover reasonable numbers of organisms from the spleen at this very early time point, a dose of 1 × 106 Listeria should be used. At 6 h postinfection, spleen Listeria CFU were already reduced by ∼1 log in mice that received the A4-8 mAb (Fig. 2 A). We interpret this result to mean that anti-LLO mAb can act to reduce Listeria growth in vivo by some very fast-acting mechanism.
To investigate whether anti-LLO Abs are generated as part of the normal immune response to Listeria infection, we performed the following experiment. CB.17 mice were infected with Listeria three times, each time with an increasing number of organisms spaced 2 to 4 wk apart. Following the third infection, serum Ab to LLO was measured in an ELISA assay with plate-bound histidine-tagged recombinant LLO (Fig. 3,B). Ab responses were minimal, with only a small fraction of the mice showing any anti-LLO Ab at a serum dilution of 1/500. To determine whether the Ab response to LLO was especially low compared with the Ab response to other Listeria Ags, we tested these same serum samples for Ab to whole Listeria Ag in a similar ELISA assay with plate-bound Listeria (Fig. 3,A). While a few mice showed detectable Ab at a serum dilution of 1/50, again very little Ab was detectable at a serum dilution of 1/500. These results indicate that Ab responses to Listeria Ags, including LLO, are very low in mice following Listeria infection. The Ig from these Listeria-immune mice did not confer passive protection to Listeria (Fig. 3 C).
In this study we demonstrate that the passive administration of a neutralizing mAb to LLO, a secreted pore-forming toxin of the facultative intracellular bacterium L. monocytogenes, can provide resistance to Listeria infection in mice. The mechanism(s) whereby Ab mediates protection is not clear at this point, although the mAb can act very quickly to limit Listeria growth in vivo. Potential mechanisms could include the following: 1) opsonization of bacteria and/or complement activation through the binding of surface associated LLO (this is unlikely since in our ELISA assay the A4-8 mAb recognized whole Listeria only at an Ab concentration of 100 μg/ml); 2) neutralization of LLO inside of the phagosomes of infected cells, leading to the prevention of bacterial escape to the cytosol; or 3) neutralization of the exotoxic functions of LLO during a time when the organism is present in the extracellular environment. Studies are in progress in our laboratory to address these possibilities.
Although Ab can clearly add a component of resistance to listeriosis, it is by no means an absolute requirement. Indeed CD4 and CD8 T cells can confer protective immunity without the need for Ab, a point exemplified in SCID mice, which clear infection after T cell transfer (20). Ab to LLO provided protection to SCID mice but did not allow for the complete clearance of the organism, indicating a necessity for T cells in mediating sterilizing immunity. Thus, we envision a cooperativity in listeriosis of different cells and soluble molecules (cytokines and Ab) to bring about complete resistance.
The surprising finding is that Ab can generate marked protection against listeriosis. Previous studies in which anti-Listeria serum was taken following infection or immunization with killed bacteria and then transferred to naive animals did not demonstrate protection against Listeria infection (2, 21, 22). Our experiments differ from previous studies with immune serum because we have utilized a defined mAb to a particular virulence factor, LLO. In infection models with other intracellular pathogens, mAbs have been shown to provide resistance when polyclonal serum could not (23). This has been explained to be due to a low abundance of serum Ab to protective Ags, or as a result of serum Ab being of an inappropriate isotype to mediate resistance.
Confirming the results of others, we found that the titer of anti-Listeria Abs, particularly anti-LLO Abs, during the normal infection of mice is limited (2, 24). This explains the failure to transfer protection with serum. Why LLO can elicit CD4 and CD8 T cell immunity but a weak B cell response may have to do with the fate of the intact LLO molecules and/or its particular biological features (25, 26). Overall, our studies demonstrate that an Ab to a secreted virulence factor of a facultative intracellular bacterium can provide resistance against infection, and open up the possibility for immunization with such virulence factors as a method for vaccination against intracellular pathogens.
We thank Katherine Frederick for technical support, Dr. Robert Schreiber for the GIR.208 mAb, Dr. Daniel Portnoy and Dr. Margaret Gedde for the recombinant LLO expression system, and Dr. Osami Kanagawa and Dr. Hao Shen for advice in many helpful discussions.
This work was supported by grants from the National Institutes of Health.
Abbreviation used in this paper: LLO, listeriolysin O.