Validation of a novel culture protocol for equine endometrial cells as a model to investigate uterine inflammation.

Implications
Uterine inflammation (endometritis) is ubiquitous in the mare post-mating but subsequent persistent mating-induced endometritis (PMIE) has implications for fertility and affects approximately 15% of Thoroughbred broodmares (Zent et al., 1998). A cell culture model was established to investigate the innate immune response of endometrial cells to inflammatory stimuli, nagting the need for whole-animal studies.

Introduction
Current methods for investigating the immune pathways that underlie PMIE and potential therapeutic targets uses the whole (live) animal. Whole animal model data accurately reflects the in vivo scenario, accounting for the interplay between internal organ and tissue systems. However, whole animal models require large numbers of animals and the application of invasive procedures such as repetitive artificially-induced endometritis and uterine biopsy. Therefore, alternative in vitro models are essential tools if synonymous of the whole animal. In vitro cell culture models are commensurate with the 3Rs of animal research; Reduction, Refinement and Replacement. Equine endometrial epithelial and stromal cells have been cultured twice previously (Watson et al., 1992; Szostek et al., 2012); however, methodological challenges were apparent. This study aimed to adapt a bovine endometrial cell culture protocol (Singh et al., 1999) and validate its use for equine endometrial epithelial and stromal cells.

Material and methods
Equine uteri were collected post-slaughter from a commercial abattoir (n=17). All uteri were in the luteal phase of the oestrous cycle. Endometrium was dissected from underlying tissue, chopped into 1mm3 sections, and plated in culture flasks. Tissue remained in each flask until a corona of mixed endometrial cells had formed. Cultures were purified via a process of differential trypsinisation. Williams complete media was used for cell culture and 0.1% and 0.25% trypsin to lift stromal and epithelial cells respectively. Once confluent; cells were challenged with control (media alone), oxytocin (OT; 1nM), or E. coli-derived lipopolysaccharide (LPS; 1µg/ml), to determine their response to physiological and immunological conditions. The response was assessed by measuring cellular prostaglandin F2α (PGF2α) and prostaglandin E2 (PGE2)¬ secretion by radioimmuno assay.

Results
Stromal cells secreted significantly more (P<0.05) more PGF2α and PGE2 when challenged with OT than control treatment. No other response to any treatments were seen by either epithelial or stromal cells. Passage number had a significant (P<0.05) effect on prostaglandin secretion, with both EC and SC secreting more PGE2 in passages four and five respectively. Stromal cells secreted more PGF2α in passage five, whereas EC had a tendency (P<0.08) to secrete most in passage two. There was a large degree of inter-mare variability (P<0.001).

Conclusion
The protocol was established, optimised and validated for the culture of equine endometrial epithelial and stromal cells. The model may be used for future studies to investigate the response of equine endometrial cells to immunological challenge and the effect of potential anti-inflammatory compounds. The protocol negates the need for live, whole animals to be used in endometritis research, and will facilitate the understanding of PMIE and potential treatment strategies for improving mare health and welfare.