Background

Three-dimensional (3D) cell culture models represent a new frontier in biomedical research, dramatically changing the experimental approaches to tissue physiology and pathology. In particular, gland organoids are considered highly relevant in vitro models, since they can reconstitute the complex interactions among different cell types, which are at the basis of the tissue physiology and are often altered in pathological conditions.

In the past years, different types of epithelial organoids have been developed to study mechanisms determining reproductive tract diseases, providing an impressive advancement in modeling the reproductive system. Epithelial organoids have been established from many different regions of the female reproductive tract; among others, endometrial organoids represent nowadays a pillar in the study of endometrial dysfunctions and infertility [1]. In particular, endometrial organoids, reproducing endometrial architecture in vitro, provide an extremely valuable tool to model different physiological conditions and analyze tissue alterations occurring in pathology, as well as the efficacy of pharmacological treatments. In fact, they can recreate 3D structures from tissues derived from both healthy donors and endometrial-related disease patients.

Several different approaches have been used to develop organoid models able to reproduce endometrial complexity [2,3]. The endometrial tissue used to develop organoids can be obtained by laparoscopy or biopsy catheter, from all stages of the menstrual cycle, from the decidua, atrophic tissues, or menstrual blood [4–7], and from both fresh and cryopreserved samples [8]. Several types of pathological tissues (e.g., carcinomas, endometriotic lesions, or biopsies from infertile women) can also easily generate organoids, thus representing novel models to study endometrial physiology and pathology in different biological conditions, including all the phases of the implantation process [4,5,9].

Two seminal reports were published in 2017, in which the generation of endometrial gland organoids was described [5,7]. Organoids in both studies were generated starting from the in vitro growth of isolated glandular fragments, which were then embedded in an artificial extracellular matrix (i.e., Matrigel^TM) in chemically defined culture media. 3D cultures displayed a genetically stable phenotype and self-renewing and self-organizing abilities, preserved responsiveness to hormone exposure, and exhibited cellular differentiation upon specific hormonal stimulation. Published endometrial organoids are organized as a single columnar epithelium with a central lumen, and are suitable for long-term cultures when grown in the presence of activators of Wnt and ERK1/2 signaling and inhibited TGFβ and BMP pathways [10,11].

Although endometrial gland organoids recapitulate the features of uterine glands, preserving spontaneously organized in vitro luminal and glandular elements, they cannot reproduce the midluteal endometrium, since they lack stromal and immune cell components and are devoid of vascular elements. This limits the information obtained on epithelial–stromal interactions, and thus their suitability for many highly specific analyses in which all these cell populations play a significant role.

To overcome these issues (mainly the lack of stromal cells within the 3D culture), more complex culture models have been developed and named assembloids. These were established by incorporating stromal cells with endometrial glandular organoids after their generation [3,12]. However, this approach is technically and experimentally challenging, namely regarding the introduction of a luminal epithelium and/or uterine natural killer cells and the development of protocols for long-term maintenance and propagation. A more advanced model of assembloids has been reported, in which endometrial epithelial and stromal cells were cultured using a matrix and air–liquid interface [13]. This model seems to better recapitulate endometrium anatomy, different cell population composition, and hormone-induced changes along the endometrial menstrual cycle. However, this model is technically challenging and difficult to develop.

In this scenario, we report a novel, simple, and cost-effective endometrial organoid model based on floating Matrigel^TM droplets, which allows the formation of organoids containing both epithelial and stromal cellular components.

Starting from glandular elements isolated from fresh or cryopreserved Pipelle biopsies from women who underwent assisted reproductive technologies, we generated free-floating Matrigel^TM organoids grown in 10% fetal bovine serum (FBS)-DMEM/F12, which we named “floating organoids,” self-organized as round or elongated structures, formed by epithelial cells integrated with stromal cells around a central lumen. Floating organoids were obtained from 100% of the biopsies analyzed [14], demonstrating a high reproducibility even with very small amounts of starting tissue or when it is derived from patients.

The comparison of this novel model with conventional organoids evidenced a morphology similar to the conventional organoid, being formed by a monolayer of epithelial cells organized as round or elongated sprouted structures surrounding an empty cavity. As observed in conventional organoids, floating organoids express E-cadherin, mucin 1, and acetyl-α-tubulin, confirmed by the epithelial origin of the cells composing the glandular structure. A significant presence of vimentin-expressing stromal cells, mostly localized in close contact with epithelial cells, was detected in floating but not in conventional organoids.

In our simplified model, we propose the use of conventional medium supplemented only with FBS, representing an easy and inexpensive way to obtain 3D cultures. Compared to the use of chemically defined media, FBS provides a more suitable environment for the growth of stromal cells, which is prevented in conventional organoid cultures by a medium that may actually counteract stromal cell growth and persistence in the 3D structure. Moreover, the floating culture growth also prevents stromal cells from the organoids from growing on the plastic interface of the Petri dish, consequently favoring their integration with the endometrial epithelium. This is a particularly relevant difference between the two types of 3D cultures, since the reduced presence of stroma cells represents a major limitation of conventional organoids, as the interaction between epithelial and stromal cells plays a central role in hormonal differentiation. In fact, the persistent presence of stromal cells promotes cell differentiation, as assessed for decidualization marker expression, making floating organoids a valuable model of differentiated organoids. On the other hand, the differentiation of floating spheroids in serum-supplemented medium, without the necessity of hormonal differentiation, leads to a reduction in stemness, and the differentiation status reduces the possibility of prolonged in vitro propagation, being limited to one or two passages.

In conclusion, here we describe a simple and rapid protocol to generate differentiated endometrial organoids, which retain both epithelial and stromal cells in the 3D structure. The interaction between these cell populations may allow the study of their reciprocal modulation, in particular during the differentiation process.