Characterization of a new mouse model of spontaneous endometriosis for preclinical studies of associated endometrial and ovarian pathologies.

Emanuele Pelosi^1,2; Bhawna Kushawaha¹; Hongyu Gao³, Peter Koopman⁴.

1. Department of Biochemistry and Molecular Biology, Indiana University School of Medicine, Indianapolis, IN, USA

2. Centre for Clinical Research, The University of Queensland, Brisbane, QLD, Australia
3. Center for Medical Genomics, Indiana University School of Medicine, Indianapolis, IN, USA

4. Institute for Molecular Bioscience, The University of Queensland, Brisbane, QLD, Australia

Abstract Text:

Endometriosis affects the lives of 190 million women worldwide. There is no cure for this debilitating condition, and most of the treatment approaches, including surgical interventions, are considered temporizing rather than curative. Endometriosis is associated with infertility and higher risk of ovarian cancer. One of the major gaps in endometriosis research is the lack of animal models, which negatively impacts translation of potential therapies to clinical care. Current experimental models for endometriosis artificially introduce endometrial cells from one animal into the peritoneal cavity of the same or a different animal through surgical intervention. These approaches often require ovariectomy and manipulation of the endogenous hormonal environment. It has been stated that interpretation of data from these models requires caution, due to the confounding effects of experimental interventions impacting clinical translation.

We have developed a new mouse model of spontaneously occurring endometriosis by ablating the Hnf1b gene in the Müllerian ducts. In humans, HNF1B variants are associated with uterine anomalies, endometriosis, as well as ovarian and endometrial cancer. In the mouse, ablation of Hnf1b resulted in Mullerian duct anomaly resembling Mayer-Rokitansky-Küster-Hauser (MRKH) syndrome, characterized by Mullerian hypoplasia and obstructive malformations. MRKH syndrome is highly associated with endometriosis, and over 60% of our mice developed ovarian cysts that resembled endometriotic lesions. To characterize these lesions, we performed spatial transcriptomic and immunohistological analyses of cysts, uterine, and ovarian samples, as well as Mass Spectrometry of serum samples.

We found that ovarian cysts contained uterine epithelium, stroma, and glands, demonstrating their endometriotic nature. In the ovaries, we saw infiltration of foamy histiocytes replacing the ovarian stroma, suggesting impairment of ovarian function and potential for neoplastic transformation. In addition, uterine samples from mutant mice showed increase in inflammatory cells and markers, and older mice showed signs of neoplastic transformation, consistent with the association of HNF1B with these diseases. Mass spectrometry analysis showed that the most represented proteins were pro-inflammatory factors.

While ovarian endometriosis is known to affect fertility and increase the risk of ovarian cancer, the molecular and cellular mechanisms involved are unclear. This novel mouse model replicates several features of human endometriosis and could potentially facilitate investigation of pathogenic mechanisms and improve research translation.