Motivation
Many of the diseases relating to placental dysfunction occur in early or mid-gestation, yet our ability to study placenta dysfunction often relies on sourcing trophoblasts from either term-placenta or trophoblasts cell lines. Alternately we rely on large or small animal models to measure placental function. Arguably, all lack physiological relevance to human placental-dysfunction. Some advances have been made in placenta-on-a chip and organoids but so far, the technology has not advanced to accommodate high throughput studies required for drug discovery. We aimed to develop and validate a 3D placenta-on-a-chip model that can be used in high -throughput system using a methodology to derive-trophoblast from human induced pluripotent stems cells (HiPSC).
Methods
A commercially available HiPSC line (ChiPS4) were cultured on a ECM-collagen scaffold and seeded into Petri dish or one lane of a 3-lane OrganoPlate (Mimetas) both pre-coated with Geltrex for 2D and 3D culture respectively. A differentiation protocol was adapted from Amita et al (1). Organoplates were rocked to induce perfusion and tube formation, and the parallel lane was coated with collagen-1. Trophoblast and pluripotent markers at the RNA (PCR, RNAseq) and protein level (immunoblot, immunohistochemistry) were investigated over differentiation days(D)0-6. The integrity of the POC barrier was assessed daily (D0-6) by adding 10kDa or 155kDa- dextran linked to a fluorescent dye to the perfusable channel and imaged (Incucyte) over 20mins.
Results
In both 2D and 3D HiPSC culture differentiation protocol induced up regulation of trophoblast markers (KRT-7, GATA3, PGF ,HLA) from D2-6 of differentiation, as determined by qRT-PCR (n=4 repeats)) and immunoblotting (n=4). In accordance, down-regulation of pluripotent markers (Nanog, POU51, TBXT) at day 2 compared to day 0. In 2D culture, immunohistochemical staining of KRT7 was absent at D0 and present on D2, whereas the reverse was observed for Nanog, clearly showing trophoblast development. The HiPSC-trophoblast differentiated in the OrganoPlate formed a hollow tube-structure from D3, against the parallel channel containing ECM-collagen. RNAseq (n=4) at D2 was used to observe the changes in clusters of genes (>20) associated with different trophoblast types. At D4 differentiation there was an increase in syncytiotrophoblast whereas down-regulation of extra-villainous trophoblasts gene-clusters. Interestingly, immunochemistry showed a defined area in close proximity to the ECM appeared as a preferential site for cell fusion and β-hCG production. The placenta-on-a-chip formed a leak tight barrier from D4 differentiation which retained the 155kDa and 10kDa-dextran in the placenta-on-a-chip. Comparison of 2D and 3D culture by RNAseq showed good similarity, but interestingly in 3D culture there were enhanced representation of total genes in GO terms and Reactome pathways associated with ECM, growth factor and interferon signalling pathways.
Conclusion
We successfully developed and characterised a 3D placenta-on-a-chip model using HiPSC derived trophoblasts in perfusable multi-chip OrganoPlates. Our placenta-on-a-chip model provides an exciting potential to replace animal studies to measure maternal-fatal barrier. Moreover, the current system provides the ability for robust high throughput studies for 40-POC per plate, in a more physiologically relevant system, than current 2D- primary trophoblasts or cell lines.