A 3-dimensional microfluidic platform for modeling human extravillous trophoblast invasion and toxicological screening

Abstract

Placental trophoblast cells invasion into the maternal uterus is an essential and complex event in the formation of the maternal-fetal interface. Commonly used two-dimensional (2D) cell invasion tools do not accurately represent the in vivo cell invasion microenvironment. Three-dimensional (3D) silicone polymer polydimethylsiloxane (PDMS) microfluidic platforms are an emerging technology in developing organ-on-a-chip models. Here, we present a placenta-on-a-chip platform that enables the evaluation of trophoblast invasion with intraluminal flow within an engineered PDMS 3D microfluidic chip. This platform reproduces key elements of the placental microenvironment, including endothelial and trophoblast cells, layered with an extracellular matrix, and incorporates dynamic medium flow while allowing for real-time monitoring, imaging, evaluation of trophoblast cell invasion, and heterocellular cell-to-cell interactions. Coupled with fluorescent cell tagging and flow cytometry, this platform also allows collection of the invasive cells. This will help our understanding of pathways that regulate trophoblast cell invasion and may prove important for toxicological screening of exposures that interfere with invasiveness in a complex organ such as the placenta.

Figure 1.

A. Silicone polymer polydimethylsiloxane (PDMS) 3D microfluidic chip. B. 3D microfluidic chip scheme depicting the following components: 1) a central compartment (red) with a central feeder line supplied by two inlet ports (a, b) connected to two outlet ports (c, d); 2) two outer channels (blue) with two outer feeder lines supplied by two inlet ports (e, f) connected to two outlet ports (g, h), and a 3) pillar barrier in between. Arrow corresponds to the directionality of medium flow. C. Close up scheme of the center of the 3D microfluidic chip with the central compartment (red), outer channels (blue, width (CW): 200 μm), and pillar barrier (width (BW): 50 μm), filled with pillars (pillar spacing (PS):3 μm).

Figure 2.

Cell adhesion in ECM coated dishes and 3D-microfluidic chips. Top Panels. Cell adhesion for HTR8/SVneo trophoblast cells (A) and human umbilical vein endothelial cells (HUVECs) (B) were compared to uncoated dishes after seeding into the gelatin, Matrigel and fibronectin coated 60 mm dishes for 2, 3, and 4 h (A, B) after seeding. Different letters denote differences among groups (a ≠ b denote P < 0.05). Bottom Panel (C). Cell adhesion for HTR8/SVneo trophoblast cells and human umbilical vein endothelial cells (HUVECs) was compared to uncoated 3D microfluidic chip after seeding into the gelatin, Matrigel and fibronectin coated chips 4 h after seeding.

Figure 3.

Microfluidic barrier permeability to FITC-dextran (25 μg/ml) without (A–E) and with cells (G–M) seeded in the 3D microfluidic chip. 3D microfluidic chip permeability after 35, 75, 95, 115, and 135 seconds (s) (A–E), and 24 h (G), 48 h (H) and 72 h (I) of constant dextran flow without cells seeded (A–E) and with cells seeded (G–M) in both, the central compartment (HUVECs) and the outer channels (HTR8/SVneo). The permeability, measured as fluorescence area in the central compartment, was quantified after constant dextran flow for 35, 75, 95, 115 and 135 s (F). 3D microfluidic chip permeability after 6 h (J), 24 h (G, K), 48 h (H, L) and 72 h (I, M) of constant dextran flow with cells seeded in the chip. Dotted line in panel G represents the outline of the central compartment. Note that cells can be observed within the outline of the central compartment (H–M), while cells are not visible in the outer channels due to the high FITC-dextran concentration.

Figure 4.

Scheme of the experimental design used in the study for cell invasion in the 3D microfluidic chip. HTR8/SVneo cells were pre-treated with folic acid (FA), or FA + TUN (tunicamycin) for 48 h and seeded into the outer channels. HUVECs were cultured for 48 h and then seeded into the central compartment. After 72 h of culture (culture conditions: fibronectin, 30 million cells/ml, and 0.01 μl/min flow speed), invading mCherry tagged HTR8/SVneo cells can be imaged under the fluorescence microscope, fixed for immunostaining, or trypsin harvested for cell sorting and further analysis, such as further culture, real-time quantitative PCR (qPCR) or western blot (WB).

Figure 5.

3D microfluidic chip immunofluorescence of HTR8/SVneo (outer channels) and HUVECs (central compartment) after 72 h of culture (fibronectin, 30 million cells/ml, and 0.01 μl/min flow speed). Bright field (A), DAPI nuclear stain (B), HUVECs immunostained for von Willebrand factor (vWF, C), mCherry-tagged HTR8/SVneo cells (D) and merged images (E).

Figure 6.

Cell invasion in the 3D microfluidic chip. The effect of control (vehicle, DPBS), folic acid (FA, 100 ng/ml) and FA (100 ng/ml) plus tunicamycin (TUN, 50 ng/ml) on mCherry-tagged HTR8/SVneo cells invasion after 24, 48 and 72 h exposure in the 3D microfluidic chip with (top 3 panels) or without HUVECs (bottom panel) seeded in the central compartment (culture conditions: fibronectin, 30 million cells/ml, and 0.01 μl/min flow speed).

Figure 7.

HTR8/SVneo cell invasion in a transwell after 18 h exposure to the control (vehicle, DPBS) (A), folic acid (FA, 100 ng/ml) (B), and FA plus tunicamycin (TUN, 50 ng/ml). Transwells were counterstained with DAPI (C), and quantified (D). Different letters denote differences among groups (a ≠ b ≠ c, denote P < 0.05).

Figure 8.

Flow cytometry and gene expression after invasion assay in the 3D microfluidic chip. After 3 days of culture (culture conditions: fibronectin, 30 million cells/ml, and 0.01 μl/min flow), cells in central compartment and outer channels were trypsin-digested and harvested for cell sorting (see Methods for details). (A) Representative flow cytometry result for Texas Red⁺ (right, mCherry tagged HTR8/SVneo cells) and Texas Red⁻ (left, HUVECs cells) of cells in the central compartment. (B) Matrix metalloproteinase-2 (MMP2) mRNA expression in sorted Texas Red⁺ cells from the central compartment and outer channels in control (C) and folic acid (FA) groups. Different letters denote differences among groups (a ≠ b denote P < 0.05).