Report

Extracellular Replication of Listeria monocytogenes in the Murine Gall Bladder

See allHide authors and affiliations

Science  06 Feb 2004:
Vol. 303, Issue 5659, pp. 851-853
DOI: 10.1126/science.1092712

Abstract

The bacterium Listeria monocytogenes can cause a life-threatening systemic illness in humans. Despite decades of progress in animal models of listeriosis, much remains unknown about the processes of infection and colonization. Here, we report that L. monocytogenes can replicate in the murine gall bladder and provide evidence that its replication there is extracellular and intraluminal. In vivo bioluminescence imaging was employed to determine the location of the infection over time in live animals, revealing strong signals from the gall bladder over a period of several days, in diseased as well as asymptomatic animals. The data suggest that L. monocytogenes may be carried in the human gall bladder.

Listeriosis remains a serious foodborne illness with a high mortality rate (1). The causative agent, Listeria monocytogenes, is a facultative intracellular, Gram-positive bacterium that can grow at low temperatures, putting many food products at risk for contamination. The infection of susceptible inbred Balb/c mice by this organism is among the most well characterized animal models of infectious disease and immunity. Bacterial virulence determinants such as listeriolysin O (LLO) and ActA have been extensively analyzed in cell culture and in animals (2, 3). Molecular details of species specificity (4) and cell-mediated immunity (5, 6) have been elucidated. Despite these advances, many key aspects of the infection are still uncharacterized, including the exact roles of several bacterial virulence determinants in the infection of various host cells and tissues.

The maintenance of several genes involved in specific pathogenic processes (7) suggests that L. monocytogenes has adapted to live in more than one environment in the host. Sequential analyses of the entire body over time may reveal a more complete picture of the overall disease process than standard methods, such as the quantitation of colony-forming units (cfu) of bacteria in organs known to be infected. Here, we report the use of in vivo bioluminescence imaging (BLI) to reveal the spatiotemporal distribution of L. monocytogenes in infected Balb/c mice and to provide evidence that the bacterium can grow extracellularly in the lumen of the gall bladder.

L. monocytogenes strain 10403S and the American Type Culture Collection L. monocytogenes strain LM23074 were made luminescent by chromosomal integration of a lux-kan transposon cassette (fig. S1A), which was modified for optimal expression in Gram-positive bacteria, as previously described for Staphylococcus aureus (8) and Streptococcus pneumoniae (9). The patterns of bioluminescence after intravenous inoculation of a bioluminescent clone (designated 2C, an insertion into flaA derived from L. monocytogenes 10403S) into 6-week-old Balb/c mice are shown in Fig. 1 and fig. S1. The spatiotemporal distribution of L. monocytogenes in this model at 1 LD50 (the dose required for 50% lethality) reveals the organism spreading systemically, infecting mainly the liver and spleen. The LD50 of strain 2C is 4 × 104, four times that of the parental 10403S strain. The reduction in virulence is likely not due to the interruption of the flaA locus, which is not important for virulence, but instead is probably due to the ectopic expression of several new proteins, as L. monocytogenes is known to be sensitive to such expression. Signals are apparent from other organs (fig. S1, C to E), and cfu can be recovered from dissected organs once signals are detectable in vivo. Several distinct insertion sites in both L. monocytogenes 10403S and L. monocytogenes LM23074 display identical results whether the labeled strains are given orally or intravenously (10).

Fig. 1.

Bioluminescence imaging of listeriosis in Balb/c mice. (A) Course of lethal intravenous infection over 5 days. Luminescent L. monocytogenes 2C at 4 × 104 cfu per mouse (1 LD50), imaged at the indicated days post-infection (upper right numbers) with an IVIS Imaging System (Xenogen Corporation, Alameda, CA). (B) Oral infection at 5 × 109 cfu per mouse, shown on days 1 to 6. The scale bar applies to all images in Fig. 1.

The parameters that affect how photons are transmitted through tissues have been extensively studied (11). Regardless of the light source, transmission depends on the type of tissue and the depth of the light source in the animal. The oxygen dependence of bacterial luciferase has been discussed relative to in vivo imaging applications (12). The presence of signal at a particular site demonstrates sufficient oxygen for the reaction to occur, but cfu determinations must be correlated with signals.

Oral infection of 6-week-old Balb/c mice with the labeled L. monocytogenes 2C strain (at 5 × 109 cfu) resulted in strong signals in the lower abdomen at 3 hours post-infection (10), which we interpret as the inoculum. The signals diminished to undetectable levels by 18 hours, suggesting that the bacteria were largely cleared overnight. After 3 to 9 days, depending on the dose, the signals reappeared in different locations in many of these animals and persisted for 2 or 3 days before waning again (Fig. 1B). During the second appearance of the signal, mice appeared asymptomatic and healthy. These results reveal the dynamics of the infection in individual mice, indicating waves of growth and clearance that would be extremely difficult to observe with assays that require tissue sampling from different animals that are sacrificed at predetermined time points. Of particular interest to us was a strong focal signal located over the lower thoracic region, which appeared 2 days before death in all mice that were intravenously, lethally infected with 1 LD50, and at various times, relative to dose and route, in sublethally infected mice. The focal thoracic signal was localized to the gall bladder by image-guided dissection (Fig. 2, A to C). The signal was usually observed for a few days, but sometimes for more than a week, and it first appeared as late as 15 days after oral inoculation. In animals displaying only the focal thoracic signal, no other organs were luminescent on dissection. Strains recovered from the gall bladder retained both oral and intravenous virulence.

Fig. 2.

Signal localization and culture of L. monocytogenes from the lumen of the gall bladder. (A) Distinctive thoracic signal observed in many mice during oral and intravenous infection. One of six sublethally, intravenously infected Balb/c mice is shown on day 3. Three of six animals displayed the characteristic signal over a 2-day period from day 3 to day 5 (not shown). (B) Ex vivo imaging of the gall bladder removed from the mouse. (C) A petri dish with bioluminescent colonies of L. monocytogenes recovered from the lumen of the gall bladder. Colony counts indicate that this mouse had 104 bacteria per μl of lumenal contents. (D) Tissue Gram stain of an infected gall bladder.

Because L. monocytogenes could be cultured from the acellular luminal contents of the gall bladder, we were interested in determining whether the bacterium could replicate extracellularly in that organ in living mice. To localize the bacteria in the gall bladder, we used image-guided histology of infected gall bladders (Fig. 2D). The bacteria were found extracellularly in the lumen of the organ, as has been previously described for the murine bile duct of infected animals (13). The bacterial cells were most often in chains, suggesting replicative forms. The methods employed in staining are capable of revealing intracellular bacteria, but none were observed. No signs of inflammation or tissue damage were observed. Because signal intensity depends on many parameters, possibly leading to a decreased ability to detect the bacteria, photon flux was correlated with cfu counts from the gall bladder lumen (fig. S2). The results show photons and cfu counts to be related, with a correlation coefficient of 0.86 (fig. S2B).

Mutants of LLO (Δhly) are greatly attenuated in mice, primarily because of their inability to replicate intracellularly (14). To examine the effect of the Δhly mutation on the ability of L. monocytogenes to grow in the gall bladder, we transduced the 2C lux insertion (15) into an L. monocytogenes Δhly 10403S strain. Logarithmic phase–dependent BLI signals from the gall bladder would be consistent with bacterial growth in that organ. Only growing L. monocytogenes emitted detectable signal (fig. S3), indicating that metabolically active bacteria were producing the signal in vivo.

The bioluminescently labeled 2CΔhly (LLO-negative) strain was then used to infect female 6-week-old Balb/c mice. Strong signals were detected from the gall bladder, beginning 3 to 4 days post-inoculation (Fig. 3, A and B). The absence of internalins A and B does not affect the ability of L. monocytogenes to grow in the gall bladder, as a Δhly ΔinlA/B mutant retained this ability (Fig. 3C). The number of Δhly cfu recovered from gall bladders of animals with strong signals (3 × 107 cfu per gall bladder) was much greater than that recovered for Δhly bacteria in the entire liver after 24 hours (14).

Fig. 3.

Replication of hly and hly inlA/B deletion mutants in the gall bladder. (A) A luminescent hly deletion of L. monocytogenes 10403S was inoculated intravenously into 6-week-old Balb/c mice at 108 cfu per mouse. The mice were then imaged on the indicated days. Two of five animals displayed the characteristic signal of gall bladder growth. One animal is shown. (B) A higher dose (109) produced the characteristic signal in all the mice. Four mice are shown on day 5 post-inoculation. An animal displaying a similar signal yielded 3 × 107 cfu from the gall bladder lumenal contents. (C) Replication of a Δhly ΔinlA/B mutant at doses of 109 (left panels, days indicated) and 5 × 108 (right panel).

Together, the growth-phase dependence of BLI signal, the high number of bacteria recovered from the gall bladder luminal contents, the persistent presence of the signal in deletion mutants of hly, and the histological localization of the bacteria in the lumen indicate that L. monocytogenes grows extracellularly in the murine gall bladder. L. monocytogenes is known to express a bile salt hydrolase that affects virulence when deleted (7). These and other adaptations may help the bacteria to grow transiently in the gall bladder, possibly escaping the immune system and gaining spread from the secretion of bile into the intestine, perhaps to reinfect the same animal or to be transmitted (13). It is unknown whether L. monocytogenes replicates in the gall bladder of asymptomatic humans; cases of listerial cholecystitis, inflammation of the gall bladder, have been reported (16), but only in patients with severe disease. If healthy individuals support growth of L. monocytogenes in the gall bladder, listeriosis could be spread unknowingly, in a manner similar to that of typhoid fever.

Supporting Online Material

www.sciencemag.org/cgi/content/full/303/5659/851/DC1

Methods and Materials

Figs. S1 to S3

References and Notes

References and Notes

View Abstract

Navigate This Article