Recapitulation of Human Embryonic Heartbeat to Promote Differentiation of Hepatic Endoderm to Hepatoblasts

Abstract

Hepatic development requires multiple sequential physicochemical environmental changes in an embryo, and human pluripotent stem cells (hPSCs) allow for the elucidation of this embryonic developmental process. However, the current in vitro methods for hPSC-hepatic differentiation, which employ various biochemical substances, produce hPSC-derived hepatocytes with less functionality than primary hepatocytes, due to a lack of physical stimuli, such as heart beating. Here, we developed a microfluidic platform that recapitulates the beating of a human embryonic heart to improve the functionality of hepatoblasts derived from hepatic endoderm (HE) in vitro. This microfluidic platform facilitates the application of multiple mechanical stretching forces, to mimic heart beating, to cultured hepatic endoderm cells to identify the optimal stimuli. Results show that stimulated HE-derived hepatoblasts increased cytochrome P450 3A (CYP3A) metabolic activity, as well as the expression of hepatoblast functional markers (albumin, cytokeratin 19 and CYP3A7), compared to unstimulated hepatoblasts. This approach of hepatic differentiation from hPSCs with the application of mechanical stimuli will facilitate improved methods for studying human embryonic liver development, as well as accurate pharmacological testing with functional liver cells.

Keywords: heart beating; hepatic endoderm; hepatoblast; human embryonic stem cells (hESC); mechanical stimulation; microfluidic device.

FIGURE 1

(A) Illustration of an early human embryo. Heart (H: red) beating confers mechanical stimulation to the surrounding cells, and hepatic endoderm (HE: blue), which differentiates to hepatoblasts, is exposed to mechanical forces. (B,C) Appearance (B) and cross-section (C) of a microfluidic device for applying a series of stretching stimulations to hepatic endoderm-like cells (HECs). (D) Cross-sectional view of the device, composed of polydimethylsiloxane (PDMS) and consists of a top layer with cell-culture chambers, middle membrane layer, and bottom layer with pressure chambers on a glass slide. (E) Photograph of our device fabricated on a glass slide (25 × 75 mm). Culture chambers are filled with red ink. This device has two sets of culture chambers in which cells are cultured under the same intensity of mechanical stimulation. (F) Photograph of a microfluidic device. Microfluidic channels and pressure chambers filled with green ink.

FIGURE 2

Pressure-drop method to generate a series of PDMS membrane displacements in a single device. (A) Membrane displacements are inversely proportional to the distance from the inlet due to a pressure drop along with a microfluidic channel. (B,C) The computer simulation for pressure drop was conducted using input pressure (18, 32, and 45 kPa) and output pressure (atmosphere). (D) Displacement measurement of the device with CCD laser displacement camera when 18, 32, and 45 kPa were applied at the inlet. The red plots indicated the conditions used for stretching stimulation. ANOVA with Tukey-Kramer test compared with all displacements of the pressure chambers at 18, 32, and 45 kPa. p-values were summarized in Supplementary Tables S1, S2. Each plot represents the mean ± standard deviation determined from four independent experiments measuring the two chambers in a single device.

FIGURE 3

Hepatoblast differentiation from human pluripotent stem cells (hPSCs) promoted by mechanical forces on hepatic endoderm-like cells (HECs). (A) Schematic diagram showing hepatic differentiation from hPSCs. ROCK inhibitor (Y27632), WNT inhibitor (CHIR, CHIR99021), PI3K inhibitor (LY, LY294002), bFGF, BMP4, and FGF10 were used for corresponding differentiation stages. (B) Flow cytometric analyses showing the proportion of CXCR4 expression in hPSCs, HECs in a dish and HECs in a microfluidic device. (C) Bioluminescent CYP3A activity assay for HECs and hepatoblasts. ANOVA with Tukey-Kramer test compared relative light units (RLU) divided by amount of proteins in cell lysates of all samples (Supplementary Figure S6 and Supplementary Table S3). *P < 0.05, ***P < 0.001 (n = 3). Max, median, minimum of three independent experiments were shown. (D,E) Immunocytochemical analyses showing the expression of ALB, CYP3A7, (D) and CK19 (E) in HECs and hepatoblasts in the indicated conditions. Nuclei were stained with DAPI. Scale bars represent 50 μm. (F) The negative controls of immunochemistry stained with only 2nd antibody labeled with Alexa 488 and Alexa 594 fluorescent dyes shown in (F). Scale bars represent 100 μm. (G–I) Quantitative single cell profiling of ALB (G), CYP3A7 (H), and CK19 (I) in HECs and hepatoblasts in the device. Fluorescence intensity was obtained from several images per each sample. ANOVA with Tukey-Kramer test compared with mean intensity. ***P < 0.001.