Bioengineering-tissue strategies to model mammalian implantation in vitro

Abstract

During mammalian implantation, complex and well-orchestrated interactions between the trophectoderm of implanting blastocysts and the maternal endometrium lead to a successful pregnancy. On the other hand, alteration in endometrium-blastocyst crosstalk often causes implantation failure, pregnancy loss, and complications that result in overall infertility. In domestic animals, this represents one of the major causes of economic losses and the understanding of the processes taking place during the early phases of implantation, in both healthy and pathological conditions, is of great importance, to enhance livestock system efficiency. Here we develop highly predictive and reproducible functional tridimensional (3D) in vitro models able to mimic the two main actors that play a key role at this developmental stage: the blastocyst and the endometrium. In particular, we generate a 3D endometrial model by co-culturing primary epithelial and stromal cells, isolated from sow uteri, onto highly porous polystyrene scaffolds. In parallel, we chemically reprogram porcine adult dermal fibroblasts and encapsulate them into micro-bioreactors to create trophoblast (TR) spheroids. Finally, we combine the generated artificial endometrium with the TR spheroids to model mammalian implantation in vitro and mimic the embryo-maternal interactions. The protocols here described allow the generation of reproducible and functional 3D models of both the maternal compartment as well as the implanting embryo, able to recreate in vitro the architecture and physiology of the two tissues in vivo. We suggest that these models can find useful applications to further elucidate early implantation mechanisms and to study the complex interactions between the maternal tissue and the developing embryos.

Keywords: 3D endometrial model; chemical reprogramming; microbioreactors; scaffolds; trophoblast spheroid.

FIGURE 1

Generation of the porcine 3D endometrial model and its characterization. (A) Schematic representation of the 3D endometrial model assembling, from uterine biopsy to endometrial epithelial cell and stromal fibroblast isolation and culture onto 2D standard plastic dishes (scale bar 100 μm). (B) Hematoxylin and Eosin staining of the 3D endometrial model obtained by co-culturing epithelial cells and stromal fibroblasts onto highly porous polystyrene scaffolds (scale bars 100 and 50 μm). (C) Transcription levels of VIM, COL3A1, KRT18, ZO1, CDH1 and EPCAM genes in endometrial stromal cells cultured on 2D culture systems (blue bars), endometrial epithelial cells cultured on 2D culture systems (pink bar), 3D endometrial models (yellow bars) and in vivo endometrial tissue, as a positive control (green bars). Gene expression is presented with the highest level set to 1 and all others relative to this. Data are expressed as the mean ± the standard error of the mean (SEM). ^a,bDifferent superscripts indicate p < 0.05. ND: not detected. (D) Immunofluorescent staining of 3D endometrial models for VIM (red) and ZO1 (green). Nuclei are counterstained with DAPI (blue) (scale bars 100 and 20 μm).

FIGURE 2

Porcine 3D endometrial model validation. (A) TEER values detected in 3D endometrial model obtained by co-culturing endometrial stromal and epithelial cells onto highly porous polystyrene scaffolds at different time points. Data are expressed as the mean ± the standard error of the mean (SEM). ^a,b,cDifferent superscripts indicate p < 0.05. (B) PGE2 release in 3D endometrial models after 24-h OT + AA stimulation (OT + AA) and in untreated cells (Control). Data are expressed as the mean ± the standard error of the mean (SEM). ^a,bDifferent superscripts indicate p < 0.05. (C) Transcription levels of COX1, COX2, PGFS and PTGES genes in 3D endometrial models after 24-h OT + AA stimulation (OT + AA) and in untreated cells (Control). Gene expression is presented with the highest value set to 1 and all others relative to this. Data are expressed as the mean ± the standard error of the mean (SEM). ^a,bDifferent superscripts indicate p < 0.05.

FIGURE 3

Generation of TR spheroids and their characterization. (A) Dermal fibroblasts grown into 2D culture systems were resuspended in medium containing 1uM 5-aza-CR and encapsulated onto PTFE micro-bioreactor. After 18 h of culture, the obtained pluripotent spheroid was recovered, encapsulated into a new micro-bioreactor and cultured for 11 days using a TR induction medium (scale bars 100 and 50 μm). (B) Transcription levels for fibroblast- (THY1 and VIM), pluripotent- (OCT4, NANOG, REX1, and SOX2) and TR-related (GCM1, PPAG3, ING, PAG6, HSD17B1) genes in untreated fibroblasts (T0, blue bars), in pluripotent spheroids (green bars) and in TR spheroids (lilac bars).Gene expression is represented with the highest value set to 1 and all others relative to this. Data are expressed as the mean ± the standard error of the mean (SEM). ^a,bDifferent superscripts indicate p < 0.05. ND: not detected. (C) Representative images of a TR spheroid immunostained for GATA3. Nuclei are counterstained with DAPI (blue) (scale bars 50 μm).

FIGURE 4

Representative image of a TR spheroid attached to the 3D endometrial model. (A) Hematoxylin and Eosin staining of a TR spheroid adhering to the 3D endometrial culture system after a 24-h co-culture. No space between the spheroid and the epithelial compartment was visible (scale bar 50 μm). (B) Immunohistochemical co-staining for the mature TR marker GATA3 (red) and ZO1 (green) of a TR spheroid adhering to the 3D endometrial culture system. Nuclei were counterstained with DAPI (blue) (scale bars 100 μm).