Endometrial Glandular Epithelial Cell Plasticity During Postpartum Regeneration And Estrous Cyclicity

Several different mechanisms of regeneration have been proposed to explain repair of the endometrial luminal epithelium (LE) following menstrual/estrous cyclicity, parturition or other damaging insults to the uterine mucosa. These include activity of an as-yet unidentified epithelial stem cell population, mesenchymal transdifferentiation, or integration of bone marrow-derived cells. Here, we used a novel Cag-Flox-STOP-Flox-Pgrmc2-3XFlag (Pgrmc2-3XFlag) reporter mouse coupled with a recently developed GE-specific cre recombinase mouse (Prss29-Cre) to determine if and to what extent the GE contributes to regeneration of the LE following parturition and during aging in nulliparous mice. To initiate regeneration in the endometrium, control (Pgrmc2-3XFlag) and reporter (Prss29-Cre;Pgrmc2-3XFlag) mice were placed with males for one pregnancy. Postpartum (pp) uterine tissues were then collected 5 days (1x5dpp), 3-weeks (1x3wpp) and 7-weeks (1x7wpp) later. A fourth set of uterine tissues was collected 3-weeks after three pregnancies (3x3wpp). The number of glands did not change at any time point. Indelibly marked FLAG+ GE cells were detected by immunohistochemistry and immunofluorescence in GE as expected. As demonstrated in serial uterine sections of uteri from all four treatment groups, GE cells migrated toward the LE and formed junctional points (JP) where GE cells transitioned from cuboidal GE cells to pseudostratified columnar LE cells. Transitioning cells at junctional points expressed markers of both GE (i.e., FOXA2) and LE (i.e., CALB1) tissues. Following transition, GE-derived LE cells expressed PGR and ESR1 consistent with surrounding developmentally-derived LE cells. About 17 FLAG+ LE cell islands were also observed in 1x5dpp 4 mm longitudinal sections at sites distant from GE-LE junctional points. Interestingly, the number of islands did not change from 1x5dpp to 1x7wpp. However, the total number of GE-derived FLAG+ LE cells observed in each island decreased from 1x5dpp to 1x3wpp, but then tended (p=0.07) to increase by 1x7wpp. There was no difference in the number of GE-LE junctional points or GE-derived FLAG+ LE islands between 1x3wpp and 3x3wpp tissues. However, a significant increase in the total (p=0.02) and average (p=0.02) number of FLAG+ LE cells was observed in 3x3wpp LE islands highlighting the proliferative expansion of these cells over the intervening ~80 days. GE-LE junctional points were also observed in 5-month old nulliparous Prss29-Cre;Pgrmc2-3XFlag mice at diestrus, but not in younger 6-week old mice. Collectively, these findings establish the GE as a quantifiable cellular source for LE regeneration during postpartum involution and as part of normal LE cell turnover during the estrous cycle in nulliparous mice. We propose a mechanism whereby GE cells within glandular structures are not terminally differentiated, but rather are stably differentiated to perform glandular function. However, when needed and based on the level of endometrial damage, GE cells can dedifferentiate, become integrated in to LE tissue and then reprogram to take on an LE phenotype as part of the regenerative process following parturition and during estrous cyclicity over time.