Metformin prevents age-associated ovarian fibrosis by modulating the immune landscape in female mice.

Ovarian fibrosis is a pathological condition associated with aging and is responsible for a variety of ovarian dysfunctions. Given the known contributions of tissue fibrosis to tumorigenesis, it is anticipated that ovarian fibrosis may contribute to ovarian cancer risk. We recently reported that diabetic postmenopausal women using metformin had ovarian collagen abundance and organization that were similar to premenopausal ovaries from nondiabetic women. In this study, we investigated the effects of aging and metformin on mouse ovarian fibrosis at a single-cell level. We discovered that metformin treatment prevented age-associated ovarian fibrosis by modulating the proportion of fibroblasts, myofibroblasts, and immune cells. Senescence-associated secretory phenotype (SASP)-producing fibroblasts increased in aged ovaries, and a unique metformin-responsive subpopulation of macrophages emerged in aged mice treated with metformin. The results demonstrate that metformin can modulate specific populations of immune cells and fibroblasts to prevent age-associated ovarian fibrosis and offers a new strategy to prevent ovarian fibrosis.

Automated classification of multiphoton microscopy images of ovarian tissue using deep learning.

Histopathological image analysis of stained tissue slides is routinely used in tumor detection and classification. However, diagnosis requires a highly trained pathologist and can thus be time-consuming, labor-intensive, and potentially risk bias. Here, we demonstrate a potential complementary approach for diagnosis. We show that multiphoton microscopy images from unstained, reproductive tissues can be robustly classified using deep learning techniques. We fine-train four pretrained convolutional neural networks using over 200 murine tissue images based on combined second-harmonic generation and two-photon excitation fluorescence contrast, to classify the tissues either as healthy or associated with high-grade serous carcinoma with over 95% sensitivity and 97% specificity. Our approach shows promise for applications involving automated disease diagnosis. It could also be readily applied to other tissues, diseases, and related classification problems.