Interferon-gamma promotes abortion due to Brucella infection in pregnant mice

Abstract

Background: The mechanisms of abortion induced by bacterial infection are largely unknown. In the present study, we investigated abortion induced by Brucella abortus, a causative agent of brucellosis and facultative intracellular pathogen, in a mouse model.

Results: High rates of abortion were observed for bacterial infection on day 4.5 of gestation, but not for other days. Regardless of whether fetuses were aborted or stayed alive, the transmission of bacteria into the fetus and bacterial replication in the placenta were observed. There was a higher degree of bacterial colonization in the placenta than in other organs and many bacteria were detected in trophoblast giant cells in the placenta. Intracellular growth-defective virB4 mutant and attenuated vaccine strain S19 did not induce abortion. In the case of abortion, around day 7.5 of gestation (period of placental development), transient induction of IFN-gamma production was observed for infection by the wild type strain, but not by the virB4 mutant and S19. Neutralization of IFN-gamma, whose production was induced by infection with B. abortus, served to prevent abortion.

Conclusion: These results indicate that abortion induced by B. abortus infection is a result of transient IFN-gamma production during the period of placental development.

Figure 1

B. abortus infection causes abortion in pregnant mice. Groups of four pregnant mice were infected with bacteria intraperitoneally at each gestation time point. On day 18.5 of gestation, their fetuses were removed. Fetuses were determined to be alive if there was a heartbeat, and dead if there was no heartbeat. Fetuses underlined were aborted.

Figure 2

B. abortus infection predominantly in placenta. Groups of four pregnant mice were infected with bacteria intraperitoneally on day 6.5 of gestation. On day 18.5 of gestation, their placenta, spleen, liver, lungs, and kidneys were removed and homogenized in saline. Tissue homogenates were serially diluted with PBS and plated on Brucella agar to count the number of CFU in each organ. Statistically significant differences bacterial numbers between placenta and spleen are indicated by asterisks (*, P < 0.01).

Figure 3

B. abortus predominantly invades into trophoblast giant cells in placenta. Placentas not used for bacterial culture were fixed in situ within uteri in 10% neutral buffered formalin and processed routinely for histologic examination and scoring using Meyer's hematoxylin stain. Specific labeling of B. abortus in placental sections was performed using the Dako EnVision System with anti-B. abortus rabbit serum and replicated bacteria are shown in brown. Normal (A and B) and infected placenta (C and D) are shown. Panels (B) and (D) indicate magnified images in panels (A) and (B). Arrows show trophoblast giant cells.

Figure 4

Transient increase in IFN-γ in pregnant mice induced by B. abortus infection. The IFN-γ (A), and TNF-α (B) serum levels were measured in each mouse by ELISA at the indicated numbers of days after infection. The mean and SE for groups of five mice are shown.

Figure 5

Preventing abortion by neutralizing IFN-γ. IFN-γ was neutralized in mice by administering anti-mouse IFN-γ monoclonal antibody (clone HB170) in vivo using 1 mg of antibody in a volume of 0.5 ml intraperitoneally 24 h before infection. Control mice were given 1 mg of normal rat IgG in 0.5 ml according to the same injection schedule as the corresponding anti-IFN-γ monoclonal antibody treatment groups. (A) IFN-γ in serum measured by ELISA. The mean and SE for groups of five mice are shown (B) Number of fetuses. Statistically significant differences between the untreated control and antibody treated mice are indicated by asterisks (*, P < 0.01).