Pathogenic outcome following experimental infection of sheep with Chlamydia abortus variant strains LLG and POS

Abstract

This study investigated the pathogenesis of two variant strains (LLG and POS) of Chlamydia abortus, in comparison to a typical wild-type strain (S26/3) which is known to be responsible for late term abortion in small ruminants. Challenge with the three strains at mid-gestation resulted in similar pregnancy outcomes, with abortion occurring in approximately 50-60% of ewes with the mean gestational lengths also being similar. However, differences were observed in the severity of placental pathology, with infection appearing milder for strain LLG, which was reflected in the lower number of organisms shed in vaginal swabs post-partum and less gross pathology and organisms present in placental smears. Results for strain POS were somewhat different than LLG with a more focal restriction of infection observed. Post-abortion antibody responses revealed prominent differences in seropositivity to the major outer membrane protein (MOMP) present in elementary body (EB) preparations under denaturing conditions, most notably with anti-LLG and anti-POS convalescent sera where there was no or reduced detection of MOMP present in EBs derived from the three strains. These results and additional analysis of whole EB and chlamydial outer membrane complex preparations suggest that there are conformational differences in MOMP for the three strains. Overall, the results suggest that gross placental pathology and clinical outcome is not indicative of bacterial colonization and the severity of infection. The results also highlight potential conformational differences in MOMP epitopes that perhaps impact on disease diagnosis and the development of new vaccines.

Fig 1. Histopathological changes in the placentas of sheep infected with C. abortus strains.

(A) S26/3, note large areas of infiltration by large numbers of leucocytes into the epithelium of the cotyledon forming large amounts of necro-suppurative material (arrows); (B) S26/3, the leucocytes were mixed, primarily neutrophilic, many degenerate, inflammatory cells; (C) S26/3, arteritis as denoted by infiltration of medium to large numbers of leucocytes into the blood vessel wall, which were (D) neutrophils and fewer monocytes; (E) S26/3, severe fibrinoid necrosis as denoted by a dense band of intensely eosinophilic material within the arterial wall (arrows); (F) S26/3, early thrombosis of blood vessels. Note large thrombus comprised of palisading layers of degenerate erythrocytes, leukocytes and proteinacious material along with loss of definition of the tunica intima (arrows). (G) LLG, note less extensive infiltration of leucocytes (arrows) with less penetration into the deeper tissue compared to S26/3. (H) POS, note less extensive infiltration of leucocytes (arrows), especially the deeper tissues beneath the epithelium, compared to both S23/6 and LLG, the lesions were more associated with the intercotyledonary membrane than cotyledonary villi but of a similar morphology being comprised of mixed leucocytes but primarily neutrophils (I). (J) Negative control, note absence of inflammatory cells and necrotic material.

Fig 2. Immunohistochemical detection of chlamydial antigen in placentas.

Sheep placentas infected with C. abortus strains S26/3 (A and B), LLG (C and D), POS (E) and negative control (F) and counterstained with haematoxylin. (A) S26/3, note extensive positive labelling and (B) that this was most intense and abundant in the trophoblast layer, both still attached and sloughed, and the immediately surrounding cells. (C) LLG, note positive labeling restricted to the epithelium (trophoblast) layer of the cotyledon, primarily at the bases of the cotyledonary villi and (D) in occasional foci in the intercotyledonary membrane. (E) POS, note positive labeling restricted to the epithelium (trophoblast) layer and immediately adjacent cells of the placenta in a multifocal distribution. Intense inflammatory cell infiltration, primarily neutrophils, was also present but not usually associated with the immunolabelling (arrows). (F) Negative control, note total absence of any immunolabelling and no significant infiltration by inflammatory cells.

Fig 3. Serological responses to C. abortus strains.

Detection of C. abortus antibody in ewes that aborted (includes non-viable births) (-—-) and lambed (-. -) following challenge with C. abortus strains S26/3 (A), LLG (B) and POS (C) or with negative control inoculum (D). Means (± SEM) of normalised responses (see Materials and Methods) are shown. 100% is equivalent to an OD450nm of 2.25. The lambing/abortion period is indicated by the horizontal double-headed arrows.

Fig 4. Immunoblot analysis of C. abortus EB preparations.

EBs prepared from C. abortus strains S26/3 (lane 1), LLG (lane 2) and POS (lane 3) were run on 4–12% x NuPAGE Bis-Tris polyacrylamide gels under reducing conditions and immunoblotted with pooled post-abortion sheep serum from ewes infected with C. abortus strains S26/3 (A), LLG (B) and POS (C) as detailed in Materials and Methods. Molecular masses (kDa) are as indicated.