Absolute uterine factor infertility (AUFI) affects 3–5% women and has multifactorial causes. Uterus transplantation is the main treatment option for AUFI1. Alternatively, artificial uterus can be used as an incubator for embryos. A bioartificial uterus requires a complete structure that includes cell components; however, poor cell sources hamper the clinical applications of bioartificial uterus.
Mammalian uterus glandular epithelium (GE) is necessary for blastocyst implantation and its deficiency can lead to infertility2. GE responds to the ovarian hormones, estrogen (E2), and progesterone (P4), and induces the expression of leukemia inhibitory factor (Lif) for the embryo implantation3,4. Direct reprogramming of the fibroblasts by specific transcription factors or chemical molecules have been previously reported5,6. However, obtaining uterine cells through fibroblasts reprogramming has rarely been reported. In this study, we reported the generation and characterization of full chemical-induced uterine GE from fibroblasts with the ability of amplification in vitro and responding to ovarian hormones. Our study identified a new cell source that can be used for the reconstruction of the uterus or other organs and suggested a potential value for treating uterine infertility and the related reproductive medical disorders.
Mouse embryonic fibroblasts (MEFs) were isolated from embryos at embryonic day 13.5 (E13.5) and then cultured with a chemical cocktail consisting of FGF2, BMP4, mLif, 1,4-DPCA (a competitive inhibitor of HIF prolyl 4-hydroxylase), A8-301 (an inhibitor of TGF-β type I receptor ALK5 kinase), and CHIR-99021 (a GSK-3α/β inhibitor)-inducing medium (FBLDAC). Briefly, MEFs were cultured in the inducing medium for 12 days, then maintained in the expanding medium (see Supplementary methods) (Fig. 1a). Finally, the FBLDAC-induced epithelial cell-like colonies were obtained (Fig. 1b). Furthermore, these colonies could be passaged for more than 20 times with a uniform colony morphology on the matrigel- or 10% FBS-coated petri dish (Fig. 1c) and a stable karyotype of “38 + XX” or “38 + XY” (Fig. 1d, Supplementary Fig. S1a). These cells expressed the specific markers of epithelial cells including KRT19, EPCAM, CDH1 and proliferative protein, KI67 (Fig. 1e). Next, we evaluated the tumorigenicity by subcutaneous transplantation of the induced cells and mouse embryonic stem cells (ESCs) in nude mice. Our results showed that there was no teratoma formation among the ten mice that were transplanted with the induced cells for 2 months, while eight teratomas formation was observed in the ten mice transplanted with mouse ESCs for 3 weeks (Fig. 1f).
To confirm that the epithelial cells were indeed induced from MEFs, we isolated MEFs from transgenic mice carrying fibroblast-specific protein 1 (Fsp1)-Cre/ROSA26mTmG to trace the fibroblasts. The fibroblasts would permanently express the membrane-targeted GFP after the Fsp1-Cre-mediated excision of membrane-targeted tomato expression (Supplementary Fig. S1b). These traced cells ruled out contamination of cells with epithelial fate (Supplementary Fig. S1c) and were induced according to the process shown in Fig. 1a. The green colonies of the induced cells emerged after the induction by FBLDAC (Supplementary Fig. S1d). The epithelial fate of the induced cells was confirmed by immunofluorescence (Supplementary Fig. S1e).
To further characterize these induced epithelial cells, we sequenced the whole transcriptome of the chemical-induced glandular epithelial cells (named as ciGEs hereinafter), MEFs and primary epithelial cells isolated from mouse uterus (priUterus). Hierarchical clustering analysis revealed that the gene expression patterns of ciGEs were closely related to priUterus, while differed significantly from MEFs (Fig. 1g). Furthermore, ciGEs have high expression levels of Sprr2f and Ltf, similar to GE7 (Fig. 1g). With Gene Ontology (GO) analysis, we found that genes expressed in ciGEs were related to the uterus development and estrogen response. We found that ciGEs were mainly enriched with GO terms associated with epithelial fate and functions when compared with MEFs (Supplementary Fig. S2) which is indicative of the uterine glandular epithelial characteristics of ciGEs. We further detected the high expression of uterine glandular epithelial specific markers (FOXA2 and SOX17) (Fig. 1h, upper panel, Supplementary Fig. S3a) and adult stem/progenitor cell markers (Fig. 1h, lower panel, Supplementary Fig. S3b) in ciGEs. Moreover, when partially digested, ciGEs formed a gland-like structure with lumen (Fig. 1i), which implies that the ciGEs can undergo adenogenesis. We also detected the upregulation of E2- and P4-responsive genes in ciGEs (Fig. 1j, Supplementary Fig. S3c and d), which suggested that the ciGEs can respond to ovarian hormones. Additionally, hormone treatment influenced the expression dynamics of E2- and P4-responsive genes, including Lif (Fig. 1k), a critical regulator expressed in the uterine glands to promote embryo implantation8. Besides, we also found that the upregulation of implantation-related genes (Fig. 1l). These results suggested an expandable and functional uterine GE fate can be induced from fibroblasts by a chemical cocktail.
In all mammals, the uterus contains glands in the endometrium and becomes receptive when the embryo is ready to be implanted. At this stage, the uterine environment is conducive to the embryo growth and is mainly regulated by E2 and P49. Uterine glands and their secretions are necessary to establish uterine receptivity and embryo implantation, early pregnancy failure is usually caused by the secretory phase defect10. In this study, we obtained ciGEs with uterine glandular epithelial characteristics and which respond to ovarian hormones, implying their application in endometrium replacement in clinical treatment. CiGEs were obtained by induction of fibroblasts using only a chemical cocktail. The chemical cocktail has several advantages including cell permeability, convenient handling, being nonimmunogenic, and ease of standardization. These advantages make it an attractive strategy for clinical application in treatment of uterine diseases like AUFI. Furthermore, we found the upregulation of functional related genes including estrogen and progesterone response genes in ciGEs. The molecular mechanism of the chemical cocktail-induced trans-differentiation needs to be explored in the future. Nevertheless, our research provides a clue to generate target cells from resident fibroblasts in situ in damaged or aging uterus through an induction by a chemical cocktail. It also provides an in vitro model for the study of embryo implantation and loss of uterine structure or function. Meanwhile, our results provide new insights into treatments for uterine factor infertility and uterine regeneration.
Brännström, M. et al. Uterus transplantation: a rapidly expanding field. Transplantation 102, 569–577 (2018).
Gu, T. P. et al. The role of Tet3 DNA dioxygenase in epigenetic reprogramming by oocytes. Nature 477, 606–610 (2011).
Carson, D. D. et al. Embryo implantation. Dev. Biol. 223, 217–237 (2000).
Demir, R. et al. Structural differentiation of human uterine luminal and glandular epithelium during early pregnancy: an ultrastructural and immunohistochemical study. Placenta 23, 672–684 (2002).
Cieslar-Pobuda, A. et al. Transdifferentiation and reprogramming: overview of the processes, their similarities and differences. Biochim. Biophys. Acta Mol. Cell Res. 1864, 1359–1369 (2017).
Xie, X., Fu, Y. & Liu, J. Chemical reprogramming and transdifferentiation. Curr. Opin. Genet. Dev. 46, 104–113 (2017).
Han, X. et al. Mapping the mouse cell atlas by microwell-seq. Cell 173, 1307 (2018).
Stewart, C. L. The role of leukemia inhibitory factor (LIF) and other cytokines in regulating implantation in mammals. Ann. N. Y. Acad. Sci. 734, 157–165 (1994).
Wang, H. & Dey, S. K. Roadmap to embryo implantation: clues from mouse models. Nat. Rev. Genet. 7, 185–199 (2006).
Filant, J. & Spencer, T. E. Uterine glands: biological roles in conceptus implantation, uterine receptivity and decidualization. Int. J. Dev. Biol. 58, 107–116 (2014).
We appreciate the discussion of all members in Dr. Zhou’s lab and Dr. Li’s lab. We thank Yufei Li, Yuhuan Li, and Guihai Feng for their assistance in experiment and data analysis. We also thank Shiwen Li and Xili Zhu for their assistance in two-photon imaging. This work was supported by the National Natural Science Foundation of China (31621004 to Q. Z.), the National Key Research and Development Program (2017YFA0103803 to Q. Z., 2018YFA0107703 to W. L., 2017YFF0204601), CAS Strategic Priority Research Program (XDA16000000 to Q. Z. and W. L.), Key Research Projects of the Frontier Science of CAS (QYZDY-SSW-SMC002 to Q. Z.), and the Ferring Institute of Reproductive Medicine, a strategic collaborative research program of Ferring Pharmaceuticals and the Chinese Academy of Sciences (FIRMD180304 to W. L.).
Conflict of interest
The authors declare that they have no conflict of interest.
The accession number for the transcriptome of each sample data reported in this paper is GSA: CRA001472.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.