Sodium Bicarbonate Enhances the Severity of Infection in Neutropenic Mice Orally Inoculated with Listeria monocytogenes EGD

Abstract

Epidemiological studies have suggested an association between antacid therapy and development of listeriosis in humans. In this study we used a neutropenic mouse model to demonstrate that oral administration of sodium bicarbonate shortly before intragastric (i.g.) inoculation with Listeria monocytogenes EGD (serotype 1/2a) significantly increased the severity of the resulting systemic infection. An explanation for this observation is provided by evidence that L. monocytogenes EGD is rapidly inactivated in synthetic gastric fluid at pH below 5. A second strain of L. monocytogenes (CM [serotype 1/2b]) exhibited little ability to cause systemic infection following i.g. inoculation and was not significantly enhanced by administration of sodium bicarbonate. Strain CM was readily inactivated in synthetic gastric fluid even at pH 7. These data suggest that gastric acidity and enzymes provide some innate defense against gastrointestinal listeriosis in neutropenic mice.

Listeria monocytogenes is a significant food-borne pathogen, causing an estimated 2,000 human cases, and as many as 500 deaths per year, in the United States (10, 18). Listeriosis is one of the most severe food-borne infections, with a relatively high percentage of cases requiring hospitalization or resulting in death (10, 18). Our understanding of the pathogenesis of listeriosis in the gastrointestinal (g.i.) tract is very limited. Although listeriosis is widely used as a model for studying the molecular pathogenesis of intracellular pathogens and for the protective immune response against the same, the overwhelming majority of these studies use parenteral (intravenous or intraperitoneal [i.p.]) challenge rather than inoculation via the g.i. tract. In the limited number of studies that did use g.i. inoculation, most required relatively large challenge doses of L. monocytogenes to cause systemic infection (2, 7, 11, 12, 16). However, a few studies reported experimental infection of mice with relatively low numbers of L. monocytogenes (8, 13), suggesting that conditions within the murine g.i. tract might vary in their influence on the pathogenesis of listeriosis.

We and others reported previously that administration of a monoclonal antibody (MAb) (RB6-8C5) that depletes mice of neutrophils (2, 4) greatly reduces their resistance to L. monocytogenes infection via the g.i. tract (5). Some investigators have reported that neutralization of gastric pH can also enhance the virulence of L. monocytogenes in the g.i. tract of experimental rodents (15, 17). In the present study, we sought to combine these two previous observations in an investigation of whether neutralization of gastric pH can further enhance the ability of L. monocytogenes to infect neutropenic mice via the g.i. tract.

L. monocytogenes strain EGD (serotype 1/2a) has been previously used and reported on by our laboratory (3, 11). Strain CM (serotype 1/2b), which was obtained from the culture collection of the Food Research Institute (Madison, Wis.), was originally isolated from a human outbreak involving ingestion of contaminated chocolate milk (14). Both strains were inoculated into brain heart infusion (BHI) broth and incubated with shaking overnight at 37°C. Following this, the bacteria were harvested by centrifugation, resuspended in BHI containing 20% glycerol, and stored at −70°C as 1-ml aliquots. Before each experiment, an aliquot was thawed, inoculated into 50 ml of BHI broth, and incubated at 37°C with shaking until mid-log phase growth was reached. To prepare the inoculum for mice, the bacterial cells were harvested by centrifugation (2,000 rpm at room temperature for 20 min), the bacterial pellet was resuspended in the original volume of phosphate-buffered saline, and appropriate dilutions were made in sterile phosphate-buffered saline to achieve the desired bacterial concentration. The actual number of CFU in the inoculum was verified by plating on blood agar.

Eight- to twelve-week-old female outbred mice (ICR strain) were obtained from Harlan Sprague Dawley (Indianapolis, Ind.) and housed under microisolator caps at the School of Veterinary Medicine animal care facility. Mice were depleted of neutrophils by i.p. injection of 150 μg of purified RB6-8C5 MAb 24 h before bacterial challenge, as described previously (4, 5). We have demonstrated previously that this regimen results in a substantial neutropenia for at least 3 to 4 days, which greatly impairs the resistance of mice to either intravenous or intragastric (i.g.) challenge with L. monocytogenes (4, 5). Food was removed from the cages 5 h prior to i.g. inoculation with L. monocytogenes. This was done to prevent delivery of the inoculum into the stomach from being physically blocked by ingested mouse chow, which might lead to aspiration of the inoculum into the lungs. To neutralize gastric acid, some mice were given 50 μl of 10% sodium bicarbonate (wt/vol in sterile water) via a small pipette placed in the oral cavity. Fifteen minutes after administration of sodium bicarbonate, mice were lightly anesthetized by i.p. injection of sodium pentobarbital (1 mg per mouse). Once mild sedation had occurred, the listerial inoculum was introduced directly into the stomach (in a total volume of 0.2 ml) via an infant feeding tube (3.5 french) attached to a 1-ml syringe.

At the desired time points, mice were humanely euthanatized by asphyxiation with CO₂, followed by cervical dislocation. The abdominal cavity was then aseptically opened, and portions of the spleen and liver were removed, weighed in sterile weigh boats, and placed in sterile tissue grinders that contained 1 ml of cold sterile saline. The spleens and livers were homogenized and diluted in saline, and appropriate dilutions were plated on blood agar. The plates were incubated at 37°C for 48 h, and the colonies were counted. The results are expressed as the mean ± standard error of the mean (SEM) log₁₀ CFU L. monocytogenes per g (wet weight). Data were analyzed using a repeated-measures analysis of variance. If a significant F value was obtained (P < 0.05), then the Tukey-Kramer test was performed to determine whether the means of treatment groups differed from controls. Statistical significance for all comparisons was set at P < 0.05. All calculations were performed using the Instat Biostatistics Package (GraphPad Software, Inc., San Diego, Calif.).

Administration of sodium bicarbonate to mice (50 μl of a 10% [wt/vol] solution) significantly increased the severity of i.g. infection with L. monocytogenes EGD (Fig. 1) in neutropenic mice. Approximately 2 log₁₀ more CFU of L. monocytogenes EGD were recovered from the spleens (P < 0.01) of neutropenic mice that received sodium bicarbonate, than from control mice, at 2 days after inoculation with 10⁴ L. monocytogenes EGD organisms (Fig. 1A). There was no difference (P > 0.05) in the number of CFU recovered from the liver at that time point. The effect of sodium bicarbonate treatment in neutropenic mice was more striking at 4 days after challenge, at which time nearly 6 log₁₀ more CFU of L. monocytogenes EGD were recovered from the spleens and livers (P < 0.01) than from those of control mice (Fig. 1B). The effect of sodium bicarbonate administration was dependent on the L. monocytogenes challenge dose. Approximately 3 log₁₀ more CFU were recovered from the spleens (P = 0.12) and livers (P < 0.05) of mice treated with sodium bicarbonate, compared with those of control mice, at 2 days after i.g. inoculation with 10⁶ L. monocytogenes EGD organisms (Fig. 2A). In contrast, no significant effect (P > 0.05) of sodium bicarbonate administration was observed in mice inoculated i.g. with 10⁸ L. monocytogenes EGD organisms (Fig. 2B). The lack of an effect of sodium bicarbonate treatment at the challenge dose of 10⁸ organisms is not surprising, because the number of CFU recovered from neutropenic mice without bicarbonate treatment approached the maximum one can recover from the spleen and liver (10⁸ to 10⁹ CFU per g).

FIG. 1.

Oral treatment with sodium bicarbonate increases the severity of infection with L. monocytogenes EGD in neutropenic mice. Mice were given 150 mg of RB6-8C5 MAb to deplete them of polymorphonuclear cells. Twenty-four hours later, some mice received 50 μl of 10% sodium bicarbonate (wt/vol) (Bicarb +) in the oral cavity. Fifteen minutes later, all mice were anesthetized by i.p. injection of sodium pentobarbital (1 mg) and then inoculated i.g. with approximately 10⁵ CFU of L. monocytogenes EGD, as indicated in the methods section. Mice were euthanatized 2 days (A) or 4 days (B) later, and the numbers of CFU of L. monocytogenes recovered from the spleen (▪) and liver (□) were determined by plate counts on blood agar. Results are expressed as the mean + SEM (error bar) log₁₀ CFU per g (wet weight; five mice per group). ∗, P < 0.01 compared to control.

FIG. 2.

The effect of sodium bicarbonate administration on g.i. listeriosis is less at higher L. monocytogenes EGD challenge doses. Mice were given 150 mg of RB6-8C5 MAb to deplete them of polymorphonuclear cells. Twenty-four hours later, some mice received 50 μl of 10% sodium bicarbonate (wt/vol) (Bicarb +) in the oral cavity. Fifteen minutes later, all mice were anesthetized by i.p. injection of sodium pentobarbital (1 mg) and then inoculated i.g. with approximately 10⁶ (A) or 10⁸ (B) CFU of L. monocytogenes EGD. Mice were euthanatized 2 days later, and the numbers of CFU of L. monocytogenes recovered from the spleen (▪) and liver (□) were determined by plate counts on blood agar. Results are expressed as the mean + SEM (error bar) log₁₀ CFU per g (wet weight; five mice per group). ∗, P < 0.05 compared to control.

The EGD strain of L. monocytogenes used in this study is serotype 1/2a. We went on to examine a second strain (serotype 1/2b) of L. monocytogenes (herein designated CM). This particular strain was isolated from an outbreak in which people developed gastrointestinal distress following ingestion of L. monocytogenes organisms that were present in chocolate milk that had been temperature abused (14). This strain exhibited a limited ability to cause systemic infection in neutropenic mice (Fig. 3); administration of sodium bicarbonate did not significantly increase the severity of systemic infection compared to that in control mice.

FIG. 3.

Effect of sodium bicarbonate administration on the severity of infection in neutropenic mice i.g. inoculated with L. monocytogenes strain CM. Some mice received 50 μl of 10% (wt/vol) sodium bicarbonate (Bicarb +) in the oral cavity. Fifteen minutes later, all mice were anesthetized by i.p. injection of sodium pentobarbital (1 mg) and then inoculated i.g. with approximately 10⁵ CFU of L. monocytogenes CM. Mice were euthanatized 2 days later, and the numbers of CFU of L. monocytogenes recovered from the spleen (▪) and liver (□) were determined by plate counts on blood agar. Results are expressed as the mean + SEM (error bar) log₁₀ CFU per g (wet weight; five mice per group).

We assessed the possibility that administration of sodium bicarbonate enhanced the virulence of L. monocytogenes EGD in the g.i. tract by protecting it against inactivation at low pH. This was tested by subjecting listeriae to a synthetic gastric fluid solution (8.3 g of proteose peptone, 3.5 g of d-glucose, 2.05 g of NaCl, 0.6 g of KH₂PO₄, 0.11 g of CaCl₂, 0.37 g of bile, 0.1 g of lysozyme, and 13.3 mg of pepsin per liter of distilled water) as described previously (3). The gastric fluid was adjusted to pH 2.5 by addition of HCl, and then various amounts of sodium bicarbonate (10% [wt/vol] solution) were added to achieve a final pH that ranged from 3.5 to 7.0. To test the effects of pH on survival of listeriae, 1 ml of log-phase L. monocytogenes (in sterile saline) was added to 9 ml of gastric fluid (prewarmed to 37°C) and incubated at 37°C. At 30-min intervals samples were removed from the tubes, diluted in 1% sterile peptone broth, and plated on blood agar. The plates were incubated at 37°C for 48 h, and the resulting number of CFU of L. monocytogenes was calculated. L. monocytogenes EGD was readily inactivated in synthetic gastric fluid at pH below pH 5.0 (Fig. 4A). L. monocytogenes strain CM was inactivated in synthetic gastric fluid even at pH 7.0, suggesting that the pepsin or lysozyme in the gastric fluid was deleterious to the organism. Because previous estimates of the pH in the murine stomach have been placed at 2.5 to 4.5 (3), these data provide an explanation of why administration of sodium bicarbonate enhances the virulence of L. monocytogenes strain EGD and why strain CM was of low virulence, in i.g.-inoculated neutropenic mice.

FIG. 4.

Survival of L. monocytogenes strain EGD (A) and strain CM (B) incubated at 37°C in synthetic gastric fluid at various pHs. Synthetic gastric fluid solution (8.3 g of proteose peptone, 3.5 g of d-glucose, 2.05 g of NaCl, 0.6 g of KH₂PO₄, 0.11 g of CaCl₂, 0.37 g of KCl, 0.05 g of bile, 0.1 g of lysozyme, and 13.3 mg of pepsin per liter of distilled water) was adjusted to pH 2.5 by addition of HCl, and then various amounts of sodium bicarbonate (10% [wt/vol] solution) were added to achieve a final pH that ranged from 2.5 to 7.0 (2.5, ▪; 3.5, •; 4.0, ▴; 5.0, □; 6.0, ▵; and 7.0, ○). To test the effects of pH on survival of listeriae, 1 ml of log-phase L. monocytogenes (in sterile saline) was added to 9 ml of gastric fluid (prewarmed to 37°C) and incubated at 37°C. At 30-min intervals samples were removed from the tubes, diluted in 1% sterile peptone broth, and plated on blood agar. The plates were incubated at 37°C for 48 h, and the resulting number of CFU of L. monocytogenes was calculated. (B) The values for pH 2.5, 3.5, and 4.0 are not illustrated because all were below the limits of detection within 15 min. L. monocytogenes CM incubated in BHI broth is included as a positive control (×). The results illustrated are the mean ± SEM of (error bar) of three (A) and two (B) separate experiments.

The results of this study indicate that the severity of listeriosis in neutropenic mice is enhanced by administration of sodium bicarbonate solution before i.g. inoculation with the widely used laboratory strain L. monocytogenes EGD. This observation is consistent with previous epidemiological evidence linking antacid use with clinical listeriosis and increased fecal shedding of L. monocytogenes (1, 18). There is also a previous report that reducing gastric acidity, by cimetidine treatment, increases the severity of gastrointestinal listeriosis in experimentally inoculated rats (17). However, cimetidine can also have immunomodulatory effects via its actions on H-2 receptors. For example, cimetidine has been shown to decrease interleukin-12 and increase interleukin-10 production by murine lymphocytes (6). Such changes in cytokine expression might be expected to impair anti-Listeria resistance (9, 19, 20), although a previous study reported no effect of cimetidine on gastrointestinal listeriosis in mice (8). Our in vitro data suggest that sodium bicarbonate might act largely by protecting L. monocytogenes EGD and CM against the inhibitory effects of low pH in the murine stomach. However, we cannot exclude the possibility that changes in pH might also affect expression of virulence determinants by L. monocytogenes. A recent report indicated that glutamate decarboxylase is critical for the resistance of L. monocytogenes to the acidity of gastric fluid, and that strain EGD has low glutamate decarboxylase activity (3). The recent elegant report by Lecuit et al. (11) provided evidence for the argument that the interaction of internalin on L. monocytogenes EGD with E-cadherin on intestinal epithelial cells is a key step in translocation of listeriae across the g.i. tract in mice. However, much higher numbers of listeriae (10⁹ to 10¹⁰ CFU) were required for invasive infection in that study than in the present one. Our data indicate that factors other than the interaction of internalin with E-cadherin can influence the pathogenesis of listeriosis in the g.i. tract.

It should be noted that the present study was performed solely with serotype 1/2 strains of L. monocytogenes, which are usually associated with sporadic cases of human listeriosis (10, 18, 21). We did not evaluate serotype 4b strains, which are usually responsible for large outbreaks of food-borne listeriosis (10, 18, 21). We have preliminary evidence (data not shown) that certain serotype 4b strains of L. monocytogenes (particularly Scott A) may be less affected by sodium bicarbonate administration before i.g. challenge than is strain EGD. The results of the present study suggest that the absence of neutrophils and reduction in gastric acidity can dramatically increase the severity of g.i. listeriosis with some strains of L. monocytogenes.