In this study, we established a novel mouse model of adenomyosis induced by direct mechanical puncture of the uterus. The established lesions had glandular tubular structures consisting of epithelium and stroma within the myometrium, similar to lesions seen in human adenomyosis. We also quantitatively assessed the changes in these lesions over time and found that the lesions decreased in number and increased in size. We also found that, over a 2-month period, epithelial cell proliferation decreased, stromal cell proliferation increased, and vascular density and fibrosis increased. In addition, we found the mice were capable of pregnancy and delivery, and that their lesions persisted after delivery.
This novel adenomyosis model was established by mechanically disrupting the endometrial-myometrial interface. As mentioned in the introduction, several theories have been proposed to explain the etiology of adenomyosis. At least one type of adenomyosis (Type 1 adenomyosis in Kishi’s classification) is known to be more prevalent in women of multiparity and women who have undergone uterine surgery [24–26]. It is widely believed that this type of adenomyosis is caused by mechanical damage to the uterus that disrupts the endometrial-myometrial interface and allows the endometrium to invade the myometrium [19] [24] [27, 28]. In the present mouse model, we attempted to disrupt the endometrial-myometrial interface with repeated punctures that penetrated the entire myometrium. Using this method, we succeeded in generating glandular structures consisting of epithelium and stroma in the myometrium, which resembled human adenomyosis.
One of the advantages of this model is that following uterine puncture, which was a short procedure, the lesion persisted for a long time. Previously described mouse models of adenomyosis take significantly longer to establish. In previous studies, lesions have only been established 10–12 weeks after intervention [20] or 6 weeks after neonatal tamoxifen administration [15]. In contrast, in our model, lesions are established by day 7 at the latest, which has the advantage of allowing us to study drug administration and pregnancy prognosis early after uterine puncture, while the mice are young. Furthermore, this model was shown to create lesions that persisted for at least 65 days. In contrast, previous studies on mouse models did not examine whether the lesions persisted for a long time, except for a single study by Shen et al. [17] where lesions were examined on day 60. Thus, this model could be useful to confirm the efficacy of long-term drug treatment.
In the current study, we also established a novel quantitative analysis method for adenomyosis lesions, which involved counting and measuring the volume of lesions using slices of the entire horn. Furthermore, by using markers for epithelial cells and myocytes, we established a method to measure and evaluate the epithelial and stromal parts of the lesion independently. This quantitative analysis allowed for a better understanding of temporal changes in lesion growth and regression. The present study revealed that, over time, the number of lesions significantly decreased, and the volume of each lesion significantly increased in both the epithelial and stromal parts. However, it is unclear whether this was due to the fusion of multiple lesions or the regression of some lesions and the growth of others. This is an area that requires further study.
Following on from the previously described finding regarding the increase in lesion size over time, the current study also examined the changes in cell proliferation for both epithelial and stromal compartments over time using Ki67 staining. We found that, in the epithelial compartment, Ki67-positive cells were significantly more abundant in the early stage. In contrast, Ki67-positive cells became significantly more abundant in the stromal compartment as the lesion progressed. This suggests that, in adenomyosis lesions, epithelial cells proliferate first and stromal cells proliferate later, but it is still unclear whether this is due to; epithelial-mesenchymal transition (EMT) or because the epithelial cells cause the attraction and the proliferation of stromal cells by some unknown mechanism. Cell proliferation in eutopic and ectopic endometrium has been studied in endometriosis [29, 30], but not in adenomyosis; therefore, further studies using human specimens and comparisons with the current study should be conducted in the future.
In addition to quantitative assessment, we also qualitatively assessed the characteristics of the lesions, including vessel density and fibrosis. Regarding the vascular density, it has been demonstrated that microvessels develop around adenomyosis lesions in human adenomyosis [31, 32]. In the present study, the vascular density progressively and significantly increased over time, which is consistent with findings made in studies using other mouse models of adenomyosis [17, 18]. It is possible that the proliferation of these microvessels was responsible for the increase in the volume of adenomyosis lesions over time in our model. It is well known that human adenomyosis lesions are accompanied by fibrosis [33] due to repeated cyclic bleeding associated with tissue injury and repair [34, 35]. In our model, the extent of fibrosis in the adenomyotic lesion significantly increased over time, a finding that is consistent with discoveries made in other studies [17]. The gross findings that the adenomyosis horns were shorter than the control horns may also be due to the contraction of the muscle layer caused by fibrosis. In summary, this model allows quantitative observation of angiogenesis and fibrosis associated with adenomyosis over time and is expected to be useful in determining the efficacy of treatment methods in the future.
Furthermore, we confirmed that these mice were able to conceive and deliver, and that the lesions persisted after delivery. To the best of our knowledge, there has been only one report [19] that studied the course of pregnancy using an adenomyosis mouse model, but that report was limited to examining litter size by sacrificing dams on their fourth day of pregnancy. Therefore, the current study is the first study in which mice were allowed to deliver in order to analyze perinatal outcomes and the status of adenomyosis after delivery. One advantage of our model over the previous model is that it does not require hormonal modification or oophorectomy, which can cause ovarian dysfunction or uterine thinning [13–19] and may affect the course of pregnancy. This study showed that the number and size of adenomyosis lesions did not change between the pre-pregnancy and post-partum period; therefore, this model may be useful for modelling perinatal outcomes in pregnancies complicated by adenomyosis.
Finally, our preliminary data on perinatal outcomes showed that adenomyosis mice seemed to have an earlier delivery date and smaller litter size than normal healthy mice of the same strain and age, as reported in previous studies [21–23]. We also found that the number of implantation sites was comparable between the adenomyosis horn and the control horn, suggesting that the presence of adenomyosis not only affects implantation but also affects intrauterine development. These outcomes seem to mimic the high incidence of preterm birth, fetal growth restriction (FGR) and intrauterine fetal death (IUFD).in pregnant women with adenomyosis [9–12], and this mouse model is expected to be a suitable model of pregnancy complicated by adenomyosis, although further validation is warranted.