Bovine herpesvirus 4 is tropic for bovine endometrial cells and modulates endocrine function

Abstract

Bovine postpartum uterine disease, metritis, affects about 40% of animals and is widely considered to have a bacterial aetiology. Although the gamma-herpesvirus bovine herpesvirus 4 (BoHV-4) has been isolated from several outbreaks of metritis or abortion, the role of viruses in endometrial pathology and the mechanisms of viral infection of uterine cells are often ignored. The objectives of the present study were to explore the interaction, tropism and outcomes of BoHV-4 challenge of endometrial stromal and epithelial cells. Endometrial stromal and epithelial cells were purified and infected with a recombinant BoHV-4 carrying an enhanced green fluorescent protein (EGFP) expression cassette to monitor the establishment of infection. BoHV-4 efficiently infected both stromal and epithelial cells, causing a strong non-apoptotic cytopathic effect, associated with robust viral replication. The crucial step for the BoHV-4 endometriotropism appeared to be after viral entry as there was enhanced transactivation of the BoHV-4 immediate early 2 gene promoter following transient transfection into the endometrial cells. Infection with BoHV-4 increased cyclooxygenase 2 protein expression and prostaglandin estradiol secretion in endometrial stromal cells, but not epithelial cells. Bovine macrophages are persistently infected with BoHV-4, and co-culture with endometrial stromal cells reactivated BoHV-4 replication in the persistently infected macrophages, suggesting a symbiotic relationship between the cells and virus. In conclusion, the present study provides evidence of cellular and molecular mechanisms, supporting the concept that BoHV-4 is a pathogen associated with uterine disease.

Fig. 1

Representative phase contrast images (10X) of pure population of bovine endometrial epithelial (a) and stromal cells (b).

Fig. 2

Representative fluorescence (green) and phase contrast images (10X) of bovine endometrial epithelial (a) and stromal (b) cells at different time (12, 24, 48, 72 h) post infection (P.I.) with 1 m.o.i. of BoHV-4 EGFPΔTK and the respective phase contrast images of uninfected control. Spreading of the infection can be observed by the green colour invading the field during the time and the CPE is morphologically appreciable by the change of the cell shape, where the cells tend to shrink, becoming roundest and detaching the flask surface. The experiment was repeated three times giving the same result. c) Assessment of cell survival measured by the reduction of 3-[4, 5-dimethylthiazol-2-yl]-2, 5-diphenyl tetrazolium bromide (MTT) at 48 h post infection. Hundred percent of survival was considered the uninfected control. Values are the mean ± S.D. of three independent experiments (n= 8 per treatment, P≤0,009).

Fig. 3

a) 1, 5 % agarose gel electrophoresis of DNA extracts from bovine endometrial epithelial, stromal and control BEK cells, infected with 5 m.o.i. of BoHV-4 and BVDV control virus, at 72 h post infection. b) Titers of BoHV-4 particles released by different cell types at 48 h post infection. Values are the mean ± S.D. of three independent experiments (n= 3 per treatment, P≤0, 05).

Fig. 4

a) BoHV-4 plaques obtained at different time (1, 2, 3, 4, 5 days) post electroporation (P.E.) of nude BoHV-4EGFPΔTK DNA in different cell types. Images are representative for bovine epithelial, stromal and BEK cells. b) Graph bars, summarizing the time necessary for plaques formation in different cell types (endometrial epithelial cells, endometrial stromal cells, BT, BEL, MDBK and BEK cells). Values are the mean ± S.D. of three independent experiments (n= 3 per treatment, P≤0, 05).

Fig. 5

a) Diagram showing the 1,143 bp IE2 promoter sequence containing the putative TATA box (underlined in black) and the first 15 non coding nucleotides of the first exon. Sense and antisense primers used for the PCR amplification are in red and contain the NdeI and NheI restriction sites (underlined in red) respectively for sub-cloning of the amplicon in front of the EGFP ORF (green box) followed by the bovine growth hormone polyadenylation signal (pA (grey box)). b) Representative fluorescence and phase contrast images of transfected endometrial stromal cells with the above described construct and expressing EGFP 24 hours post transfection. c) Summarising schema of the time necessary for EGFP accumulation into the different cell types (endometrial epithelial cells, endometrial stromal cells, BT, BEL, MDBK and BEK cells), following transfection with the above described construct. Values are the mean of three independent experiments (n= 3 and P≤0, 05).

Fig.6

Prostaglandin production by (a) epithelial and (b) stromal cells after challenge with LPS (1 μg/ml) or BoHV-4 at the different m.o.i. indicated. After 24 h in culture, supernatants were harvested and prostaglandin production was measured by RIA. Values are presented as the mean ± S.D. of three independent experiments. Differences were statistically different at P < 0.05 compared with a media control. c) Western immunoblotting of endometrial stromal cells extracts and infected with BoHV-4 EGFPΔTK. Positive control was established treating cells with medium containing 20% of FBS. Beta actin as the loading control. The same experiment was repeated three times, giving the same results.

Fig. 7

a) Schematic illustration of the transwell model used for co-culturing BoHV-4EGFPΔTK persistently infected bovine macrophage cell line (in green in the upper well) producing low amount of infectious virus (grey dots) with bovine endometrial stromal or control BEK cells (in yellow) seated in the lower well. The transwell consists of a permeable membrane and allows diffusion of the viral particles. b) Fluorescence and phase contrast images of bovine endometrial stromal and control BEK cells seated in the lower well at 48 hours (48h) post co-cultivation. c) Graph bar of the viral titer in the upper well. Values are the mean ± S.D. of three independent experiments (n= 3 and P≤0,05).