Developing a Cost-Effective and Scalable Production of Human Hepatic Competent Endoderm from Size-Controlled Pluripotent Stem Cell Aggregates

Abstract

Dynamic suspension culture of human pluripotent stem cells (hPSCs) in stirred bioreactors provides a valuable scalable culture platform for integrated differentiation toward different lineages for potential research and therapeutic applications. However, current protocols for scalable and integrated differentiation of hPSCs limited due to high cost of growth factors and technical challenges. Here, hPSCs aggregates primed with 6 and 12 μM of CHIR99021 (CHIR), a Wnt agonist, in combination with different concentrations of high cost Activin A (10, 25, 50, 100 ng/mL). We sought to determine the appropriate treatment duration for efficient and cost-effective differentiation protocol for foregut definitive endoderm production in a dynamic suspension culture. Afterward, we evaluated the impact of the initial hPSC aggregate sizes (small: 86 ± 18 μm; medium: 142 ± 32 μm; large: 214 ± 34 μm) as critical bioprocess parameter on differentiation efficacy at the beginning of induction. The results indicated that 1-day priming of hPSCs as 3D aggregates (hPSpheres) with 6 μM CHIR followed by treatment with a low concentration of Activin (10 ng/mL) for 2 days resulted in efficient differentiation to definitive endoderm. This finding confirmed by the presence of ≥70% SOX17/FOXA2-double positive cells that highly expressed the anterior endodermal marker HEX. These endodermal cells differentiated efficiently into mature functional hepatocytes [60% albumin (ALB)-positive cells]. The results showed that the initial size of hPSC aggregates significantly impacted on the efficacy of differentiation. The medium sized-hPSpheres resulted in higher productivity and differentiation efficiency for scalable hepatocytes production, whereas small aggregates resulted in significant cell-loss after CHIR treatment and large aggregates had less efficacious endodermal differentiation. Differentiated cells exhibited multiple characteristics of primary hepatocytes as evidenced by expressions of liver-specific markers, indocyanine green and low-density lipoprotein uptake, and glycogen storage. Thus, this platform could be employed for scalable production of hPSC-derived hepatocytes for clinical and drug discovery applications.

Introduction

Robust, cost-effective and scalable culture system for large-scale production of hepatocyte-like cells (HLCs) from human pluripotent stem cells (hPSCs) will facilitate potential, widespread applications of these valuable cells in drug screening, liver tissue/organ engineering, and cell-based therapies. To date, a number of research groups have developed protocols for large scale expansion of hPSCs as 3D aggregates (hPSpheres) in stirred suspension bioreactors under batch or perfusion culture modes that maintain pluripotency and genetic stability as the most critical quality parameters [1]. However, current protocols for scalable and integrated differentiation of hPSCs toward HLCs suffer from critical challenges such as the high cost of growth factors, low differentiation efficacy and/or productivity, heterogonous hepatosphere formation, poor functionality of generated hepatocytes, and protocol reproducibility [2].

An efficient, cost-effective protocol for integrated differentiation of hPSCs toward functional hepatocytes can boost the potential applications of hPSC-derived hepatocytes [3]. Here, we have focused on increasing the cost-effectiveness of our previously established protocol for hepatic differentiation of hPSpheres in a dynamic suspension culture [2] and optimized the initial aggregate size as a critical bioprocess parameter to start the integrated differentiation process form hPSCs to definitive endoderm and HLCs [4].

To date, the majority of endoderm differentiation protocols have been conducted as 2D/monolayer cultures that use high concentrations of Activin [5] in the presence or absence of Wnt signaling pathway inducers [6 –8]. However, these protocols are not convenient for scalable and integrated differentiation of hPSCs since they depend on high amounts of expensive growth factors and have not been tested for hPSCs aggregate differentiation. Although researchers attempted to replace Activin with small molecules such as IDE1/2 [9] or LY294002 [10], these modifications resulted in low reproducibility and efficiency for different hPSC lines and were not applicable for integrated differentiation of 3D aggregates under suspension culture conditions [2,11].

In the current study, we primed hPSC aggregates with various concentrations of CHIR99021 (CHIR), a Wnt agonist small molecule, in combination with different concentrations of high cost Activin growth factor. We assessed treatment duration to ascertain an efficient, cost-effective differentiation protocol to foregut definitive endoderm and hepatocyte production under the dynamic suspension culture. We evaluated our 3D integrated differentiation procedure in comparison with an efficient 2D protocol established in our previous studies. In all of our reports, we found that the 3D differentiation of hPSCs, resulted in HLCs with higher associated gene expression, secretion of ALB and metabolic activities than 2D culture [12,13].

Moreover, different studies demonstrated that hPSCs colony and aggregate size as well as their heterogeneity significantly influenced differentiation trajectory and efficacy [4,14,15]. Numerous reports explored the effect of aggregate size on hPSC directed and integrated differentiation efficacy and outcome, and employed different strategies to generate size-controlled/defined size aggregates such as AggreWell™ plates and differentiation to cardiac [14,16], retinal [17], and neural cell [18] lineages. However, to the best of our knowledge, there is no study that has reported optimization of hPSC aggregate size for efficient directed differentiation of hPSpheres to endodermal cells and HLCs under dynamic suspension culture conditions. Thus, developing a cost-effective protocol and exploring the optimum aggregate size for hPSCs hepatic integrated differentiation is a crucial step need to be taken before developing a robust bioprocess for large scale manufacture of functional HLCs.

Materials and Methods

hPSC aggregate culture

Here, we used hESC lines (RH5 and RH6, passages: 25–50) [19] and the hiPSC4 line (passages: 40–45) [20]. We developed a suspension culture of hPSpheres in spinner flasks as previously described [21]. After two or three passages in a static suspension culture (low attachment culture dish), we transferred 1 × 10⁷ dispersed cells to 50 mL of human foreskin fibroblast-conditioned medium supplemented with bFGF (100 ng/mL; Royan Biotech) in a 100 mL spinner flask with a 40 rpm agitation rate [1]. All hPSCs culture procedures were performed under standard culture conditions of 37°C and 5% CO₂ with ∼95% relative humidity.

Supplementary Table S1(Supplementary Data are available online at www.liebertpub.com/scd) lists the details for all study-related materials.

Integrated differentiation of hPSpheres

The 4–5 day-old hPSpheres were washed in phosphate-buffered saline plus Ca²⁺ and Mg²⁺, and then cultured in basal medium for differentiation to endodermal cells (endospheres). The basal medium consisted of RPMI 1640 (Life Technologies), 1 × B-27 without vitamin A (-vitA) or insulin (-Ins; Life Technologies), and 0.1% bovine serum albumin (BSA; Life Technologies). On the first day, CHIR99021 (CHIR; Stemgent) was added to the basal medium. After 24 h, the cells were washed before changing the medium. Then, Activin A (R&D Systems; hereafter Activin) was added for 1–4 days to induce the hPSCs into a definitive endoderm. These differentiated endodermal aggregates were called endospheres.

The endospheres were induced for hepatic differentiation as previously described [22] in DMEM/F12 supplemented with 2% knockout serum replacement (KOSR; Life Technologies), fibroblast growth factor 4 (FGF4, 10 ng/mL; Royan Biotech) and hepatocyte growth factor (HGF, 10 ng/mL; R&D Systems) for 6 days. Next, the cells were treated for an additional 12 days in the same medium plus hepatocyte complete media without EGF (HCM, 50% v/v; Lonza), oncostatin M (OSM, 10 ng/mL; R&D Systems), and dexamethasone (Dex, 10⁻⁷ M; Sigma Aldrich).

Sphere size and morphology evaluation

The morphology and the size distribution of spheres in each step were assessed by a phase-contrast microscope (IX51; Olympus). The diameter of spheres was measured by Image J software (National Institutes of Health) under phase-contrast microscope equipped with a DP70 camera (Olympus). We counted ∼500–1,500 spheres per sample and then induced the small (86 ± 18 μm), medium (142 ± 32 μm), and large (214 ± 34 μm) hPSpheres for differentiation.

Gene expression analysis

Total RNA was extracted using TRIzol (Sigma-Aldrich) for the different groups. After removal of genomic DNA with DNaseI, we synthesized cDNA by using 2 μg total RNA with random hexamer, oligo dT, and reverse transcriptase based on the manufacturer's instructions (Takara). Real-time PCR reactions were performed as described [22]. The fold change for each gene was normalized to the housekeeping gene GAPDH and calibrated with pluripotent status. The relative gene expression levels were quantified using the 2^−ΔΔct method. Human adult and fetal liver tissues and the Huh7 cell line were used as positive controls. Supplementary Table S2 lists the primer sequences for qRT-PCR.

Immunofluorescence staining and flow cytometry

We collected, washed, and fixed the spheres overnight in 4% paraformaldehyde at 4°C for immunofluorescence staining. The fixed spheres were incorporated in an agar gel (2%). After processing, they were embedded in paraffin blocks and sectioned into 6 μm sections with a microtome (Microm™, HM325). We followed the standard protocol for immunofluorescence staining.

For flow cytometry analysis, we dissociated the differentiated spheres into single cells using trypsin/EDTA. Dispersed cells were fixed in 4% paraformaldehyde at 4°C for 20 min, permeabilized, blocked in serum, and allowed to incubate overnight with diluted primary antibodies at 4°C. Then, they were washed and incubated with secondary antibodies for 45 min at room temperature. Flow cytometry analysis was performed with a BD FACSCalibur flow cytometer (Becton Dickinson).

Supplementary Table S3 lists the antibodies used in this study.

Functional analysis

We performed periodic acid-Schiff (PAS) staining for glycogen storage, indocyanine green (ICG; Sigma-Aldrich), and low-density lipoprotein (LDL) uptake (Biomedical Technology) of the hepatospheres to assess their functional activity [22]. Secreted albumin (ALB) in the supernatant media was assessed by enzyme-linked immunosorbent assay according to the manufacturer's instructions. Urea production was evaluated according to a colorimetric assay as previously described, and data were normalized to total protein in each group [22]. To assess the cytochrome P450 activity, the hepatospheres were divided into two groups: those treated with Rifampicin (20 μM) or 0.1% DMSO (control) for 72 h. Assessment of functional activity was performed using the P450-Assay Kit (Promega) according to the manufacturer's instructions. The CYP3A4 activity of hepatospheres was measured using a luminometer. Data were normalized to total protein in each group.

Statistical analysis

Data were presented as mean ± standard deviation (SD), and analyzed according to one-way ANOVA and Tukey post hoc analysis for significant differences between groups. The mean difference was considered statistically significant at P < 0.05.

Results

CHIR priming facilitated the differentiation of hPSpheres to endospheres at a low concentration of Activin

Initially, we sought to develop a cost-effective and efficient method to differentiate hPSpheres into competent definitive endoderm in a suspension culture. Initially the cells were primed with CHIR (12 μM) for 1 day to reduce the concentration of Activin (Fig. 1A). Flow cytometry analysis of T expression as a mesoendoderm (ME) marker after 1 day showed 67% ± 15% positive cells in CHIR-treated spheres (MEspheres; Fig. 1B).

FIG. 1.

Optimization of CHIR and Activin treatment on differentiation of hPSpheres to endospheres. Schematic picture of the differentiation protocol to determine the appropriate concentration and time of Activin treatment (A). After 1 day of treatment with 12 μM CHIR99021 (CHIR), ∼70% of the cells were T-positive (B). Flow cytometry data showed that very few CHIR-treated spheres (mesoendoderm sphere; MEspheres) in basal medium without Activin were SOX17/CXCR4 double-positive. However, in the other groups treated with Activin more than 60% of cells were SOX17/CXCR4 double-positive (C). MEspheres incubated for 1 day in basal medium without any induction (rest phase) followed by treatment with Activin had significantly reduced SOX17/FOXA2 double-positive cells (D). No significant difference existed between the B-27 supplements without vitamin A (-VitA) or without insulin (-Ins) for differentiation to endospheres (D). Data showed similar endodermal marker expressions in the developed and conventional methods that used 100 ng/mL of Activin. The expression of NANOG as a pluripotency marker significantly downregulated compared to the conventional method (E). Gene expression data showed after 1 day of CHIR priming, expressions of T and MIXL1 significantly increased. Addition of Activin to the medium resulted in reduced expressions of these genes. In contrast, expressions of SOX17, FOXA2, and CXCR4 rapidly increased, particularly after a 2-day treatment period of mesoendodermal cells with Activin. The increased expressions persisted during later days (F). Flow data of spheres at 2, 3 or 4 days after treatment with Activin showed that ∼70% of cells were SOX17-CXCR4 or SOX17-FOXA2 positive (G). This finding was confirmed by immunofluorescence images for SOX17 or FOXA2 at day 3 (H). Data are presented as mean ± SD (n = 3–5). Data were analyzed using ANOVA followed by the post hoc Tukey's tests, *P < 0.05, ***P < 0.001. hPSphere, human pluripotent stem cell sphere; Endosphere, endodermal sphere; C12, CHIR99021 (12 μM); Act100, Activin 100 ng/mL; Act10, Activin 10 ng/mL; B-27 (-VitA), B-27 supplement without vitamin A; B-27 (-Ins), B-27 supplement without insulin. Insets show the nuclei that were counterstained with DAPI. Color images are available online at www.liebertpub.com/scd

In the next step, the cells were primed for 1 day with CHIR (C12) and subsequently treated for 4 days with different concentrations of Activin (10, 25, 50, and 100 ng/mL; Fig. 1C). Flow cytometry analysis demonstrated significant enhancement (>60%) of SOX17/CXCR4 double-positive cells after treatment with Activin compared to the control (no Activin) or without priming of the CHIR groups (Fig. 1C, P < 0.001). Under this condition, the different concentrations of Activin did not affect co-expression of SOX17/CXCR4 (Fig. 1C).

After CHIR priming and a plus/minus 1-day rest (no additional treatment) in basal medium (Fig. 1A), the cells were cultured for 4 days in the presence of Activin and B-27 supplement without vitamin A or insulin (Fig. 1D). Flow cytometry analysis indicated that more than 60% of the cells co-expressed SOX17/FOXA2 without the rest condition (Fig. 1D, P < 0.05 vs. rest). However, there was no significant difference in the presence of B-27 minus vitamin A or insulin (Fig. 1D). qRT-PCR analysis showed similar results for expressions of T, MESP1, and SOX1 (data not shown). Therefore, we continued our experiments with CHIR and Activin (10 ng/mL) in the presence of B-27 supplementation without vitamin A.

qRT-PCR analysis did not show any significant differences in the expressions of SOX17 and FOXA2 in CHIR-Activin (10 ng/mL) compared with only Activin (100 ng/mL). However, NANOG and OCT4 had significantly lower expressions with CHIR-Activin treatment (P < 0.05, Fig. 1E and Supplementary Fig. S1A) that might be related to the effect of Activin on maintenance of pluripotency [23].

Then, we sought to determine the optimal period for Activin (10 ng/mL) treatment to enable efficient differentiation to an endoderm fate. We analyzed the expression kinetics of a number of key genes within 5 days of induction. Data showed that after 1 day of priming with CHIR, expressions of ME markers (T and MIXL1) increased significantly (P < 0.001, Fig. 1F). The expressions of these genes downregulated following Activin treatment. In contrast, the expressions of endodermal markers CXCR4, SOX17, and FOXA2 showed significant, rapid upregulation (P < 0.05, Fig. 1F). Flow cytometry analysis revealed no differences between the percentages of SOX17/CXCR4 and SOX17/FOXA2 double-positive cells during 2–4 days of treatment of MEspheres with Activin (Fig. 1G, Supplementary Fig. S1B). Figure 1H shows immunofluorescence images of endospheres at day 3 that expressed SOX17 and FOXA2. Our data also revealed that the newly developed method in the 3D culture system was not efficient for differentiation of hPSCs under 2D conditions (Supplementary Fig. S1C). Therefore, 1 day priming with CHIR and 2 days of 10 ng/mL Activin treatment resulted in efficient differentiation to endodermal cells in the hPSC suspension culture.

After optimizing the concentration and duration of treatment with Activin, we assessed the impact of different concentrations of CHIR (1, 3, 6, and 12 μM) for differentiation to competent endodermal cells. hPSpheres were incubated for 1 day with different concentrations of CHIR followed by 2 days of treatment with Activin (10 ng/mL; Fig. 2A). We observed that the endospheres from these four groups had a similar morphology (Fig. 2B). Gene expression analysis showed that the treatment with 1 μM CHIR followed by Activin resulted in significant reduction in SOX17 and FOXA2 expressions (P < 0.05, Fig. 2C). Next, we assessed protein expressions at the three different concentrations. Immunofluorescent images revealed that pretreatment of cells with 6 μM and 12 μM of CHIR followed by Activin generated more SOX17- or FOXA2-positive cells compared to pretreatment with 3 μM of CHIR (Fig. 2D). Flow data showed ∼70% SOX17 positive cells at both the 6 μM and 12 μM concentrations of CHIR followed by Activin treatment (Fig. 2E).

FIG. 2.

Optimization of CHIR concentration priming for differentiation into foregut endodermal cells. Schematic picture of differentiation protocol to determine the appropriate concentration of CHIR (A). hPSpheres were treated for 1 day with four different concentrations of CHIR (1, 3, 6, and 12 μM) followed by treatment for 2 days with Activin (10 ng/mL). Endospheres in these four groups had a similar morphology (B). Gene expression analysis showed that the expressions of SOX17 and FOXA2 significantly reduced at lower concentrations of CHIR (C1) followed by Activin (C). Immunofluorescence showed that pretreatment of cells with C12 and C6 followed by Activin resulted in numerous SOX17- or FOXA2-positive cells, whereas there were few SOX17- or FOXA2-positive cells when pretreated with C3 (D). Flow cytometry data showed ∼70% SOX17-positive cells in those treated with C12- and C6-Activin (E). We also observed that priming with C12 and C6 affected the anterior-posterior patterning of endospheres. C12 priming of the endospheres resulted in significantly increased CDX2 expression as a posterior marker. The expression of HEX as an anterior marker did not significantly change in the CHIR priming group (F). Immunofluorescent data for SOX17 or FOXA2-positive cells for hESC (RH6) and hiPS (iPS4) in the suspension culture indicated that this method was reproducible for other cell lines (G). Data are presented as mean ± SD (n = 3–5). Data were analyzed using ANOVA followed by the post hoc Tukey's test. *P < 0.05, **P < 0.01. C12, CHIR99021 (12 μM); C6, CHIR99021 (6 μM); C3, CHIR99021 (3 μM); C1, CHIR99021 (1 μM). Insets show the nuclei that were counterstained with DAPI. Color images available online at www.liebertpub.com/scd

However, pretreatment of hPSpheres with 6 μM and 12 μM CHIR resulted in anterior-posterior patterning of the endospheres. CDX2 expression, as a posterior endodermal marker, significantly upregulated in the cells treated with 12 μM CHIR (P < 0.01). Expression of HEX, as an anterior endodermal marker, did not significantly change in the CHIR priming group (P < 0.01, Fig. 2F). These results demonstrated that CHIR priming generated anterior-posterior committed endodermal cells in a concentration dependent manner. We observed similar results with differentiation of another hESC (RH6) line and one hiPSC (hiPSC4) line (Fig. 2G). Therefore, 1-day priming of hPSpheres with 6 μM CHIR followed by 2 days of treatment with 10 ng/mL of Activin resulted in a cost-effective and efficient differentiation protocol into foregut endospheres. Therefore, we continued our research on hESCs (RH5).

We sought to assess whether endospheres could undergo further differentiation to hepatocytes by incubating the endospheres in hepatic medium for 18 days (Fig. 3A). Assessment of the major markers of the three developmental stages showed the expressions of specific markers in each stage (Supplementary Fig. S2A). More than 80% of the hepatoblast step expressed AFP (Supplementary Fig. S2B). On day 21, we observed various sizes of hepatospheres with smooth to dentate margins (Fig. 3B). Gene expression analysis showed upregulation of AFP, ALB, CYP3A7, CYP3A4, TDO, and G6PC in HLCs, which were similar to human fetal liver or adult liver (Fig. 3C). There was less AFP expression in the hepatospheres compared to the hepatoblast stage (Fig. 3D). The majority of cells within the spheres co-expressed ALB, E-cadherin (E-cad), and CYP3A4 (Fig. 3D). E-cad expressed in the surrounding cells that covered the surface of the hepatospheres (Fig. 3D). Flow cytometry analysis of data showed that 64% ± 19% of the cells expressed ALB.

FIG. 3.

CHIR (6 μM)-primed endodermal cells had the capability for efficient differentiation into hepatocytes. Schematic picture of protocol for differentiation into hepatocytes (A). The morphology of the endospheres changed to various sized round or ridged hepatospheres (B). These hepatospheres expressed high levels of hepatospecific genes (n = 3) compared to Huh7 cells, and fetal and adult liver tissues (C). There were some area for AFP-positive cells in hepatospheres, but the majority of cells within the spheres expressed ALB (∼60%) and E-cad. Hepatospheres also expressed CYP3A4 at the protein level (D). These hepatospheres showed liver performance functions of LDL and ICG uptake, and glycogen storage (E). Hepatospheres produced ALB and urea, and responded to Rifampicin in the cytochrome p450 activity test compared to hPspheres (F). Hepatosphere, Hepatocyte sphere; HLC, Hepatocyte-like cells; ALB, Albumin; E-cad, E-cadherin; LDL, low-density lipoprotein; PAS, periodic acid schiff; ICG, indocyanine green; RLU, relative light units; N.D, not determined. Color images available online at www.liebertpub.com/scd

The hepatospheres demonstrated hepatic functions of LDL and ICG uptake, and glycogen storage (Fig. 3E). They also produced ALB and urea, and responded to Rifampicin in the cytochrome p450 activity test (Fig. 3F). Therefore, 6 μM CHIR and 10 ng/mL Activin-induced endodermal cells had the capability to differentiate to mature and functional HLCs under suspension culture conditions.

Medium sized-hPSpheres were suitable candidates for efficient differentiation into endoderm and subsequent differentiation into hepatocytes

It has been well demonstrated that homogeneity and a defined size for hPSpheres are important factors for efficient differentiation into specific lineages [14]. We have employed shear stress under dynamic and scalable suspension cultures, and culturing day for generation of size-controlled hPSpheres, which offer significant advantage in scale up uniform generation of size-controlled hPSpheres [1]. In the current study, we observed that their diameter gradually increased by day-7 postseeding (d.p.s; Fig. 4A). The diameter size of hPSpheres 2–3 days after seeding was 86 ± 18 μm, which increased to 142 ± 32 μm at 4–5 days, and 214 ± 34 μm at 6–7 days (Fig. 4A, B).

FIG. 4.

Differentiation of various sized-hPSpheres into endodermal cells. The hPSpheres gradually increased in diameter during 7 days of culture. At 2–3 days after seeding, they were ∼86 ± 18 μm in diameter (small), after 4–5 days they were 142 ± 32 μm (medium), and after 6–7 days postseeding, they were 214 ± 34 μm (large). We measured the diameters of 500–1,500 spheres to calculate the sizes in each group (A). During expansion, the hPSpheres had a round, compact morphology (B). When differentiated into endodermal cells, the endosphere sizes compared to the starting hPSpheres in the three groups did not significantly change (C, D). The total population of differentiated cells at day 3 in relation to input hPSCs increased significantly in a size-dependent manner (E). Approximately 70% of cells in endospheres from the small and medium groups were SOX17-positive cells, whereas 30% were SOX17-positive or FOXA2-positive cells in the large group (E, F). Data are presented as mean ± SD. Data were analyzed using ANOVA followed by the post hoc Tukey's test. **P < 0.01, ***P < 0.001. d.p.s, days postseeding; S, endosphere derived from differentiation of small sized-hPSpheres; M, endosphere derived from differentiation of medium sized-hPSpheres; L, endosphere derived from differentiation of large sized-hPSpheres. Insets show the nuclei that were counterstained with DAPI. Color images available online at www.liebertpub.com/scd

To elucidate the appropriate size of hPSpheres for directed differentiation, we used these 3 sizes of hPSpheres and differentiated them to endodermal cells according to the aforementioned established protocol. After 3 days, there were no significant changes in the mean sizes of the endospheres compared to the corresponding starting hPSpheres in the three groups (Fig. 4C, D). The total cell population of differentiated cells at day 3 increased in a size dependent manner compared to the starting seeding spheres. The ratio of total cell population at day 3 for small sized-hPSpheres was 1.6 ± 0.8-fold, whereas for medium-sized cells it was 5.4 ± 2.3-fold (P < 0.01), and 11.1 ± 0.1-fold for large cells (P < 0.001, Fig. 4E). However, flow cytometry analysis and immunofluorescence staining showed more SOX17 and FOXA2-positive cells in those derived from medium-sized hPSpheres in comparison with large-sized hPSpheres (P < 0.001, Fig. 4E, F). These results indicated that the size of hPSpheres could affect total cell population and the efficiency of endodermal cells. Therefore, we continued our study with medium-sized hPSpheres.

Next, we assessed the hepatic competency of the endodermal cells derived from medium sized-hPSpheres. The generated hepatospheres showed HLC gene expression pattern compare to fetal and adult livers (Fig. 5A). According to flow cytometry analysis, 60% ± 9% of the medium sized-hPSpheres were ALB-positive (Fig. 5B).

FIG. 5.

Differentiation of medium sized-hPSpheres into hepatospheres. Endospheres were differentiated to hepatocyte-like cells. qRT-PCR analysis for gene expression of hepatic markers compared to fetal and adult liver tissues (A) and flow cytometry analysis for ALB-positive cells (B) in hepatospheres. The frequency of cystic hepatospheres was 27% (C). Size distribution of cystic and dense hepatospheres showed that cystic hepatospheres ranged in diameter from 20 to 1,000 μm, but the majority (∼65%) had large diameters of >600 μm. In contrast, dense hepatospheres ranged in size from 50 to 1,000 μm (∼89%, <600 μm), (D). Morphology of cystic and dense hepatospheres (E). Gene expression analysis showed no significant differences in the expressions of hepatospecific genes in different morphology (cystic and dense) and hepatosphere sizes (F). Data are presented as mean ± SD. M, Hepatospheres derived from differentiation of medium sized-hPSpheres. Color images available online at www.liebertpub.com/scd

Based on morphology, we observed two populations of hepatospheres, transparent/cystic and solid/dense (Fig. 5C). We manually sorted the cystic and dense hepatospheres to determine which group contained more hepatocytes in hepatospheres (Fig. 5D, E). Cystic hepatospheres ranged in diameter from 20 to 1,000 μm. Approximately 65% were larger than 600 μm (Fig. 5D). However, the dense hepatospheres ranged from 50 to 1,000 μm, with ∼89% that were less than 600 μm in diameter (Fig. 5D). We sorted the dense hepatospheres derived from the medium sized-hPSpheres into three groups: <300 μm, 300–600 μm, and >600 μm (Fig. 5E). The data showed no significant differences in gene expressions among various dense and cystic hepatospheres (Fig. 5F). Immunofluorescence data also indicated that ALB, E-cad and CYP3A4 expressed well in both types of hepatospheres (Fig. 6). The hepatospheres were PAS positive and showed glycogen storage (Fig. 6). Therefore, the hepatocytes were dispersed within various hepatospheres derived from medium sized-hPSpheres.

FIG. 6.

Immunofluorescence and PAS staining for medium sized-hPSpheres into hepatospheres. Immunofluorescence data showed ALB, E-cad and CYP3A4 expressed in all hepatospheres derived from medium sized-hPSpheres (A). The hepatospheres had the ability to store glycogen in their cytoplasms (B). <300: dense hepatospheres derived from medium sized-hPSpheres with <300 μm in diameter; 300–600: dense hepatospheres derived from medium sized-hPSpheres with 300–600 μm in diameter; >600: dense hepatospheres derived from medium sized-hPSpheres with >600 μm in diameter. E-cad, E-cadherin; PAS, periodic acid-Schiff. Color images available online at www.liebertpub.com/scd

Discussion

The current study has shown that 1 day priming of hPSpheres with 6 μM CHIR followed by 2 days of treatment with 10 ng/mL of Activin results in a cost-effective and efficient protocol for differentiation into foregut endodermal cells under dynamic suspension culture conditions. This protocol offers significant advantages over the conventional protocol to induce endodermal cells from hPSCs. The conventional protocol uses 10 times more Activin (100 ng/mL) along with initial activation of the Wnt signaling pathway [6 –8].

Our previous study showed that priming with rapamycin and 4 day treatment with 50 ng/mL of Activin resulted in 70% SOX17-positive cells under 2D adherent culture condition [11]. However, this protocol was not efficient for integrated differentiation as a 3D aggregate suspension culture and generated about 40% SOX17-positive cells [2]. Recently, it was also reported that with CHIR priming it is possible to reduce the amount of Activin concentration to 50 ng/mL under 2D culture condition [24]. Our previous study showed that pretreatment of hPSpheres with a high concentration of CHIR (12 μM) for 1 day led to the efficient production of mesoendodermal cells (70% T-positive cells) in a suspension culture [16]. These cells had the potential to differentiate to beating cardiac cells [16,25] or endodermal cells [24] following treatment with either cardiac inducers or Activin, respectively. In this study, we found that treating hPSpheres with a low concentration of Activin (10 ng/mL) could adequately induce them into endodermal cells after CHIR priming, which is also reported by other studies [26]. Our data showed that a 1 day rest period for the mesoendodermal cells (without additional treatment in basal medium) led to significant reduction in SOX17/FOXA2 double-positive cell numbers. It has been also reported that incubation of mesoendodermal cells for 1 day in basal medium (rest phase) promoted efficient differentiation of mesoendodermal cells into mesodermal cells [16,25].

Additionally, we demonstrated that priming of hPSCs with a moderate concentration of CHIR (6 μM) followed by 2 days of treatment with Activin (10 ng/mL) enabled efficient generation of foregut committed endodermal cells that highly expressed HEX, which agreed with other studies [27]. In other study, a moderate concentration of CHIR (4–6 μM) has been recommended to prime hPSCs to mesoderm lineage cells that include differentiation to cardiomyocytes [28], endothelial cells [29], and kidney cells [30]. It was reported that 3 μM of CHIR with Activin efficiently differentiated hPSCs to endodermal cells [31]. Naujok et al. observed that priming with a higher concentration of CHIR (7.5 μM) reduced the expressions of endodermal markers in a 2D culture of hPSCs [24].

B-27 supplement without insulin or vitamin A have been used to develop a defined and serum-free medium for differentiation of hPSCs to endodermal cells [32,33]. We observed no differences in both type of B-27 supplements during differentiation to endodermal cells. Since insulin protects cell viability and vitamin A acts as a posteriorizing factor [34], we have used the B-27 supplement with insulin and without vitamin A in our study.

Another important challenge for efficient scale-up differentiation of hPSCs under dynamic suspension culture condition is the generation of homogenous and size-controlled hPSC aggregates. Cell-cell interactions affect the sphere's fate and cell population homogeneity in addition to controlling the diffusion rates of gases, nutrients, and metabolites within aggregates which significantly impact cell proliferation and differentiation [35]. In our previous report, we did not explore the effective size of starting hPSpheres in differentiation to hepatic endoderm cells [2]. Our results demonstrated that using small sized-hPSpheres will lead to a significant cell loss during treatment with CHIR due to combining effect of CHIR molecule toxicity and shear stress under the suspension dynamic culture condition. This will result in decreasing small aggregates mechanical integrity, releasing single cell into culture medium, and subsequently significant cell loss under dynamic suspension culture mode.

The differentiation efficacy of hPSCs to endodermal cells was higher in medium sized-hPSpheres compared to the larger hPSpheres. This might be related to lower stem cell homogeneity in larger hPSpheres, along with lower diffusion rates of differentiation factors and metabolites during the differentiation process. These result indicated that initial aggregate size not only affected differentiation efficacy, but could also largely affect the productivity of the integrated differentiation process due to cell loss during the endoderm generation stage in small-size hPSpheres.

The effect of aggregate size on spontaneous and directed differentiation has been reported by other studies [14,18]. Pettinato et al. observed that very small embryoid bodies (EBs) did not survive during differentiation and large EBs underwent core necrosis [36]. Fonoudi et al. demonstrated that large hPSpheres could not efficiently differentiate to beating cardiac spheres in spinner flask culture system [16]. To date, different approaches have been employed to generate homogenous controlled size spheres such as the hanging drop culture [37], microwell plates [36], and encapsulation technology [38,39]. However, hydrodynamic pressure generated by rotary orbital suspension culture or in stirred culture vessels and spinner flasks could be considered as more practical and scalable solution to generate uniform aggregates compared to static culture condition strategies [1,39].

Our hepatospheres cultured in the spinner flask had a more rounded shape, and were dense and homogenous in size and shape. Approximately 27% od aggregates were transparent with a large cavity surrounded by one or two cell layers. Interestingly, the cystic spheres in our developed protocols were also expressed the same hepatic markers as the dense hepatospheres. Our previous study showed that 50% of hepatospheres in the spinner flask had a cystic morphology that expressed more early hepatic markers compared to dense hepatospheres [2]. This could be related to higher efficiency, and the protocol for endodermal production and improvement of HLC induction with FGF4 and HGF [40].

Taken together, in the current study, we improved the scalable differentiation of hPSpheres to mature and functional hepatocytes by the use of cost-effective and size-controlled conditions in a spinner flask. This approach resulted in efficiently differentiated hPSpheres to 70% SOX17-FOXA2 double positive cells and 60% ALB-positive cells compared to 40% endodermal cells and 25% ALB-positive cells out of the total population as reported in the previous study [2]. This method could facilitate future large-scale production of hepatocytes for tissue engineering, drug screening assays, and liver cell therapy.

Despite the advantages of the stirred culture vessels with limited culture working volumes (15–100 mL) over static suspension culture systems for integrated differentiation of hPSCs, there could be limited diffusion of oxygen and nutrients/metabolites inside the spheres at larger working volumes. This problem could be addressed by culturing under fully controlled conditions in stirred bioreactors. Recent evidence suggested that the functions of primary rat hepatocytes in a controlled bioreactor improved compared to the uncontrolled condition [41]. Consequently, large-scale differentiation of hPSCs under controlled bioreactor conditions might increase functional hepatocyte differentiation.

Footnotes

Acknowledgments

This study was funded by Royan Institute, the Iranian Council of Stem Cell Research and Technology, the Iran National Science Foundation (INSF), and Iran Science Elites Federation to H.B. We express our appreciation to Hassan Ansari, Payam Taheri, and Fazel Sahraneshin Samani at Royan Institute for their technical assistance.

Author Disclosure Statement

The authors declare they have no competing financial interests.

References

Abbasalizadeh

, Larijani

, Samadian

and Baharvand

. (2012). Bioprocess development for mass production of size-controlled human pluripotent stem cell aggregates in stirred suspension bioreactor. Tissue Eng Part C Methods, 18:831–851.

Vosough

, Omidinia

, Kadivar

, Shokrgozar

, Pournasr

, Aghdami

and Baharvand

. (2013). Generation of functional hepatocyte-like cells from human pluripotent stem cells in a scalable suspension culture. Stem Cells Dev, 22:2693–2705.

Serra

, Brito

, Correia

and Alves

. (2012). Process engineering of human pluripotent stem cells for clinical application. Trends Biotechnol, 30:350–359.

Hookway

, Butts

, Lee

, Tang

and McDevitt

. (2016). Aggregate formation and suspension culture of human pluripotent stem cells and differentiated progeny. Methods, 101:11–20.

Cai

, Zhao

, Liu

, Ye

, Song

, Qin

, Meng

, Chen

, Zhou

, et al. (2007). Directed differentiation of human embryonic stem cells into functional hepatic cells. Hepatology, 45:1229–1239.

Hay

, Fletcher

, Payne

, Terrace

, Gallagher

, Snoeys

, Black

, Wojtacha

, Samuel

, et al. (2008). Highly efficient differentiation of hESCs to functional hepatic endoderm requires ActivinA and Wnt3a signaling. Proc Natl Acad Sci U S A, 105:12301–12306.

D'Amour

, Bang

, Eliazer

, Kelly

, Agulnick

, Smart

, Moorman

, Kroon

, Carpenter

and Baetge

. (2006). Production of pancreatic hormone-expressing endocrine cells from human embryonic stem cells. Nat Biotechnol, 24:1392–1401.

Nakanishi

, Kurisaki

, Hayashi

, Warashina

, Ishiura

, Kusuda-Furue

and Asashima

. (2008). Directed induction of anterior and posterior primitive streak by Wnt from embryonic stem cells cultured in a chemically defined serum-free medium. FASEB J, 23:114–122.

Borowiak

, Maehr

, Chen

, Tang

, Fox

, Schreiber

and Melton

. (2009). Small molecules efficiently direct endodermal differentiation of mouse and human embryonic stem cells. Cell Stem Cell, 4:348–358.

10.

McLean

, D'Amour

, Jones

, Krishnamoorthy

, Kulik

, Reynolds

, Sheppard

, Liu

, Xu

, Baetge

and Dalton

. (2007). Activin a efficiently specifies definitive endoderm from human embryonic stem cells only when phosphatidylinositol 3-kinase signaling is suppressed. Stem Cells, 25:29–38.

11.

Tahamtani

, Azarnia

, Farrokhi

, Sharifi-Zarchi

, Aghdami

and Baharvand

. (2013). Treatment of human embryonic stem cells with different combinations of priming and inducing factors toward definitive endoderm. Stem Cells Dev, 22:1419–1432.

12.

Farzaneh

, Pournasr

, Ebrahimi

, Aghdami

and Baharvand

. (2010). Enhanced functions of human embryonic stem cell-derived hepatocyte-like cells on three-dimensional nanofibrillar surfaces. Stem Cell Rev, 6:601–610.

13.

Baharvand

, Hashemi

, Kazemi Ashtiani

and Farrokhi

. (2006). Differentiation of human embryonic stem cells into hepatocytes in 2D and 3D culture systems in vitro. Int J Dev Biol, 50:645–652.

14.

Bauwens

, Peerani

, Niebruegge

, Woodhouse

, Kumacheva

, Husain

and Zandstra

. (2008). Control of human embryonic stem cell colony and aggregate size heterogeneity influences differentiation trajectories. Stem Cells, 26:2300–2310.

15.

Rahman

, Brauer

, Ho

, Usenko

, Tewary

, Zuniga-Pflucker

and Zandstra

. (2017). Engineering the haemogenic niche mitigates endogenous inhibitory signals and controls pluripotent stem cell-derived blood emergence. Nat Commun, 8:15380.

16.

Fonoudi

, Ansari

, Abbasalizadeh

, Larijani

, Kiani

, Hashemizadeh

, Zarchi

, Bosman

, Blue

, et al. (2015). A universal and robust integrated platform for the scalable production of human cardiomyocytes from pluripotent stem cells. Stem Cells Transl Med, 4:1482–1494.

17.

Yanai

, Laver

, Joe

, Viringipurampeer

, Wang

, Gregory-Evans

and Gregory-Evans

. (2013). Differentiation of human embryonic stem cells using size-controlled embryoid bodies and negative cell selection in the production of photoreceptor precursor cells. Tissue Eng Part C Methods, 19:755–764.

18.

Miranda

, Fernandes

, Pascoal

, Haupt

, Brustle

, Cabral

and Diogo

. (2015). Spatial and temporal control of cell aggregation efficiently directs human pluripotent stem cells towards neural commitment. Biotechnol J, 10:1612–1624.

19.

Baharvand

, Ashtiani

, Taee

, Massumi

, Valojerdi

, Yazdi

, Moradi

and Farrokhi

. (2006). Generation of new human embryonic stem cell lines with diploid and triploid karyotypes. Dev Growth Differ, 48:117–128.

20.

Totonchi

, Taei

, Seifinejad

, Tabebordbar

, Rassouli

, Farrokhi

, Gourabi

, Aghdami

, Hosseini-Salekdeh

and Baharvand

. (2010). Feeder- and serum-free establishment and expansion of human induced pluripotent stem cells. Int J Dev Biol, 54:877–886.

21.

Baharvand

, Larijani

and Yousefi

. (2012). Protocol for expansion of undifferentiated human embryonic and pluripotent stem cells in suspension. Methods Mol Biol, 873:217–226.

22.

Farzaneh

, Pakzad

, Vosough

, Pournasr

and Baharvand

. (2014). Differentiation of human embryonic stem cells to hepatocyte-like cells on a new developed xeno-free extracellular matrix. Histochem Cell Biol, 142:217–226.

23.

Vallier

, Mendjan

, Brown

, Chng

, Teo

, Smithers

, Trotter

, Cho

, Martinez

, et al. (2009). Activin/Nodal signalling maintains pluripotency by controlling Nanog expression. Development, 136:1339–1349.

24.

Naujok

, Diekmann

and Lenzen

. (2014). The generation of definitive endoderm from human embryonic stem cells is initially independent from activin A but requires canonical Wnt-signaling. Stem Cell Rev, 10:480–493.

25.

Lian

, Hsiao

, Wilson

, Zhu

, Hazeltine

, Azarin

, Raval

, Zhang

, Kamp

and Palecek

. (2012). Robust cardiomyocyte differentiation from human pluripotent stem cells via temporal modulation of canonical Wnt signaling. Proc Natl Acad Sci U S A, 109:E1848–E1857.

26.

Sumi

, Tsuneyoshi

, Nakatsuji

and Suemori

. (2008). Defining early lineage specification of human embryonic stem cells by the orchestrated balance of canonical Wnt/beta-catenin, Activin/Nodal and BMP signaling. Development, 135:2969–2979.

27.

Matsuno

, Mae

, Okada

, Nakamura

, Watanabe

, Toyoda

, Uchida

and Osafune

. (2016). Redefining definitive endoderm subtypes by robust induction of human induced pluripotent stem cells. Differentiation, 92:281–290.

28.

Burridge

, Matsa

, Shukla

, Lin

, Churko

, Ebert

, Lan

, Diecke

, Huber

, et al. (2014). Chemically defined generation of human cardiomyocytes. Nat Methods, 11:855–860.

29.

Bao

, Lian

, Dunn

, Shi

, Han

, Qian

, Bhute

, Canfield

and Palecek

. (2015). Chemically-defined albumin-free differentiation of human pluripotent stem cells to endothelial progenitor cells. Stem Cell Res, 15:122–129.

30.

Lam

, Freedman

, Morizane

, Lerou

, Valerius

and Bonventre

. (2014). Rapid and efficient differentiation of human pluripotent stem cells into intermediate mesoderm that forms tubules expressing kidney proximal tubular markers. J Am Soc Nephrol, 25:1211–1225.

31.

Kunisada

, Tsubooka-Yamazoe

, Shoji

and Hosoya

. (2012). Small molecules induce efficient differentiation into insulin-producing cells from human induced pluripotent stem cells. Stem Cell Res, 8:274–284.

32.

Kearns

, Genga

, Ziller

, Kapinas

, Peters

, Brehm

, Meissner

and Maehr

. (2013). Generation of organized anterior foregut epithelia from pluripotent stem cells using small molecules. Stem Cell Res, 11:1003–1012.

33.

Siller

, Greenhough

, Naumovska

and Sullivan

. (2015). Small-molecule-driven hepatocyte differentiation of human pluripotent stem cells. Stem Cell Reports, 4:939–952.

34.

McCracken

, Cata

, Crawford

, Sinagoga

, Schumacher

, Rockich

, Tsai

, Mayhew

, Spence

, Zavros

and Wells

. (2014). Modelling human development and disease in pluripotent stem-cell-derived gastric organoids. Nature, 516:400–404.

35.

Antoni

, Burckel

, Josset

and Noel

. (2015). Three-dimensional cell culture: a breakthrough in vivo. Int J Mol Sci, 16:5517–5527.

36.

Pettinato

, Wen

and Zhang

. (2014). Formation of well-defined embryoid bodies from dissociated human induced pluripotent stem cells using microfabricated cell-repellent microwell arrays. Sci Rep, 4:7402.

37.

Kurosawa

. (2007). Methods for inducing embryoid body formation: in vitro differentiation system of embryonic stem cells. J Biosci Bioeng, 103:389–398.

38.

Kerscher

, Turnbull

, Hodge

, Kim

, Seliktar

, Easley

, Costa

and Lipke

. (2016). Direct hydrogel encapsulation of pluripotent stem cells enables ontomimetic differentiation and growth of engineered human heart tissues. Biomaterials, 83:383–395.

39.

Bratt-Leal

, Carpenedo

and McDevitt

. (2009). Engineering the embryoid body microenvironment to direct embryonic stem cell differentiation. Biotechnol Prog, 25:43–51.

40.

, He

, Huang

, Xu

, Tang

, He

and Feng

. (2013). Induced maturation of hepatic progenitor cells in vitro. Braz J Med Biol Res, 46:559–566.

41.

Miranda

, Leite

, Muller-Vieira

, Rodrigues

, Carrondo

and Alves

. (2009). Towards an extended functional hepatocyte in vitro culture. Tissue Eng Part C Methods, 15:157–167.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.39 MB

0.11 MB

0.02 MB