Assessing Risk for Listeriosis
To predict the risk of human illness from exposure to L. monocytogenes, we generally rely on animal models to establish dose estimates for adverse effects. Several animal models, including mouse, guinea pig, and non-human primates, have been proposed for use in determining the dose response curve for listeriosis. We have evaluated all three models, but all have their advantages and disadvantages.
The Mongolian gerbil was recently proposed as the most appropriate small animal model for listeriosis because it is susceptible to the same InA- and InB-mediated invasion pathways known to occur in humans. We wanted to determine whether gerbils would be a more sensitive animal model and provide better dose response data for humans.
We investigated invasion and adverse pregnancy outcomes in gerbils orally exposed to L. monocytogenes, compared dose-response data to data from other animal models, and investigated differences in responses of pregnant versus non-pregnant gerbils. Pregnant gerbils were orally challenged with 0 (control), 103, 105, 107, or 109 CFU L. monocytogenes strain 12443 in sterilized whipping cream. Age-matched, non-pregnant gerbils were dosed to enable comparisons of the effect of pregnancy on susceptibility to listeriosis. All animals were sacrificed 7 days after treatment.
Pregnancy did not affect fecal shedding of L. monocytogenes. Neither pregnant nor non-pregnant groups shed L. monocytogenes after challenge with 103 CFU. However, both pregnant and non-pregnant gerbils treated with 107 or 109 CFU, respectively, shed L. monocytogenes at about the same extent for 7 days.
Pregnancy did affect invasion of L. monocytogenes into maternal tissues. At doses below 105 CFU, L. monocytogenes was not isolated from tissues of pregnant and non-pregnant gerbils. At the doses of 107 and 109 CFU L. monocytogenes, tissues from pregnant gerbils were invaded to a much greater extent (>2x) than were tissues from non-pregnant gerbils.
These studies indicate that L. monocytogenes can cause adverse outcomes in pregnant Mongolian gerbils, as fetal deaths were seen in the highest dose group (109 CFU). However, a dose-response curve for fetal mortality could not be drawn because of the lack of deaths in the lower dose groups. Thus, the LD50 falls somewhere between 107 and 109 CFU, and there is a threshold for lethality.
Results also show that the gerbil model is not more sensitive, and may actually be less sensitive, to L. monocytogenes than the guinea pig and non-human primate models of listeriosis for both invasion and adverse pregnancy outcomes. The theory that the gerbil, with InlA- and InlB-mediated pathways, would be more sensitive to fetoplacental invasion of L. monocytogenes is not supported by our study and results suggest that these pathways alone are insufficient to explain the differences between susceptibility among various animal models.
Traditionally, risk assessment has used the most sensitive animal model to predict human risk, unless mechanisms are known to be different in humans and the animal model. In the case of L. monocytogenes, the non-human primate dose response is the most sensitive, followed by the guinea pig and finally the gerbil. While the gerbil may be a useful model in learning more about the InA- and InB-mediated pathways, our study suggests there are other pathways that may play an important role in listeriosis during pregnancy.