Development of three dimensional culture and expression of integrin heterodimers in human embryonic stem cells.

Extensive effort has been made to develop a three-dimensional (3D) system for the culture of pluripotent stem cells in human and model animals, which yields lots of benefits for monitoring cell-to-cell or cell-to-environment interaction and for suggesting alternative materials for clinical cases. Initial study using animal model moved toward supporting embryonic stem cells (ESCs) self-renewal in a synthetic scaffold conjugated with suitable peptide motifs. As results, the feeder-free, 3D acellular niche consisting of vinyl sulfone (VS)-functionalized polyethylene glycol (PEG)-based hydrogel binding with extracellular matrix analogs could support ESC self-renewal, but main stemness signals were switched in the 3D environment. We employed this PEG-based hydrogel for 3D culture of human ESCs and further adjustment of hydrogel constituent made it possible to support self-renewal of three ESC lines. In this study, we examined transcriptional and translational activity of integrin heterodimers for optimizing the 3D system by using peptide motifs and subsequently elucidated that transcription and translation of integrin α 5β 1, α 6β 1 and αVβ 5 were stronger than other heterodimers in a referenced human ESC line.

Development of 3D Culture for Mouse Embryonic Stem Cells

Maintaining of self-renewal activity under designed in vitro-condition is the baseline to clinically scientifically utilize pluripotent stem cells. To date, two-dimensional (2D) system using liquid medium and plastic culture container has been conventionally employed, but it apparently lacks the critical factors for maintaining cell hemostasis likely to in vivo.- Implementation of cellular components, non-cellular environment, extracellular matrix (ECM) proteins and miscellaneous substances to culture environment, which consists of both cellular and acellular niche, is absolutely necessary for optimizing culture system. A three-dimensional (3D) system containing all necessary factors for cell hemostasis was subsequently suggested. This artificial system mainly consists of synthetic polymer and peptide motifs, which provides sufficient space for cell growth with reducing mechanical stress and stimulates cell signal for cellular activity, respectively. Compared with 2D system, this 3D system mimics in vivo-condition, which cannot only monitors cell-to-cell or cell-to-environment interaction but also suggests alternative strategy for stem cell self-renewal.

Based on this rationale, we first developed a 3D acellular niche for self-renewing of mouse embryonic stem cells (ESCs). Polyethylene glycol (PEG)-based, hydrogel with oligopeptide crosslinker was employed. In this hydrogel, vinyl sulfone (VS) functionalized to the end of PEG was attached to the −SH cysteine terminal of the crosslinker, which enables to induce cell-friendly conjugation between PEG-VS and crosslinker by Michael type addition reaction., To optimize this acellular niche, combinatorial incorporation of several peptide motifs were induced for stimulating cell activity. In previous study, we employed integrin-related peptide motifs for enhancing self-renewal of ESCs under feeder-free condition. In mouse studies, the hydrogel became more active for the stem cell self-renewal when the peptide motifs activating integrin α5β1, αvβ5, α6β1 and α9β1 heterodimers were conjugated to the hydrogel than any other combinations. Interestingly, main self-renewal signal was changed under the 3D system optimized; leukemia inhibitory factor (LIF)-related Stat3 signaling no longer acted as a limiting factor, while combinatory conjugation of integrins α5β1, αvβ5, α6β1 and α9β1 with the 3D scaffold greatly increased Akt1 and Smad 1/5/8 activation.

Application to Human ESCs

These promising results encouraged us to subsequently apply this 3D system for culturing of human ESCs. We evaluated whether referenced human ESCs could self-renew in PEG-based hydrogel of different physical properties (gel concentration and multi-arm number) in the absence of peptide motifs. Regardless of gel concentration, the 8-arm PEG-based hydrogel was the most effective (Fig. 1). Especially, the 10% (w/v) 8-arm PEG-based hydrogel could support self-renewal of three human ESC lines (H1, H9 and novo).

Figure 1. Optimization of the polyethylene glycol (PEG)-based hydrogel for the culture of human embryonic stem cells (ESCs). Cell proliferation and expression of stemness-related genes in H9 human embryonic stem cells (ESCs) was monitored. The PEG-hydrogel of 7.5%, 10%, 12.5% or 15% consisting of vinyl sulfone(VS)-functionalized multiarm PEG derivatives including 3-, 4- or 8-arm with di-cystyl peptide, were used. Clump size and gene expression were analyzed on day 9 of culture. (A) Regardless of gel concentration, large H9 ESC clumps were found in 4-arm and 8-arm hydrogels. The H9 human ESCs cultured in 3-arm hydrogel, which is softer than the 4-arm and 8-arm hydrogels, were not maintained. (B) No significant difference except for TERT was detected between 4-arm and 8-arm hydrogel groups, while only TERT expression was higher in H9 human ESCs cultured in 10 to 15% hydrogel of 8-arm than those cultured in the hydrogel of other compositions. All data shown are mean ± SD from the values of three replicates. N/A, not analyzed. All figures were reproduced with permission from Lee et al.

To further optimize this hydrogel, patterning of ECM analogs is necessary and we subsequently planned to employ integrin-related peptide motifs for stimulating self-renewing. To select suitable peptide motifs, expression of integrin heterodimers functionalized on human ESC plasma membrane was screened. H9 human ESCs were employed as a referenced human ESC line for the analysis. Both gene expression and protein translation of each integrin α and β subunit were monitored in this series of experiment.

Expression of Integrin Subunits

In separate quantification of total 25 integrin α and β subunits, higher expression were observed in integrin α5, α6 and α7 (Fig. 2A, top panel) and integrin β1 and β5 (Fig. 2A, bottom panel). Weak expression was detected in integrin α1, α2, α3, α4, α8, α9, α10, α11, αL, αM, αV and αX (Fig. 2A, top panel) and integrin β2, β3, β4, β6 and β7 (Fig. 2A, bottom panel). No transcription of integrin αE and αD (Fig. 2A, top panel) and integrin β8 (Fig. 2A, bottom panel) was detected. Protein translation of integrin α and β subuints showing increased transcription was subsequently evaluated. As shown in Figure 2B, consistency in transcription and translational levels of integrin α5 and α6 and integrin β1 were detected. However, percentage of hESCs expressing integrin β5 remained low level, and in spite of weak expression of integrin αV, strong translation of integrin αV was detected. Strong expression of integrin heterodimer conjugating integrin αV with integrin β5 was observed in hESCs. Strong integrin α7 transcription did not result in increased translation. Translational levels of integrin α1, α2 and α4 was significantly lower than those of integrin α5, α6 and αV, integrin β1 and integrin αVβ5.

Figure 2. Expression of integrin subunits in undifferentiated human embryonic stem cells (hESCs). (A) Transcriptional levels of integrin α and β subunit genes in the hESCs. Among total 17 integrin α (A, top panel) and 8 integrin β subunit (A, bottom panel) genes, significant increase of transcriptional level was detected in integrin α5, α6, α7, β1 and β5 subunit genes. (B) Translational levels of integrin α and β subunit genes in the hESCs by FACS analysis. More than 89% of hESCs expressed integrin α5, α6, αV, β1 and αVβ5. All data shown are mean ± SE of three independent experiments. a-dp < 0.0001. ND, not detected.

Increased expression and translation of integrin α5β1, α6β1 and αVβ5 was functionally validated by using of antibodies and cell culture plate coated with ECM proteins. Inhibition of integrin function against each ECM protein was induced by blocking antibodies against integrin subunits α5 or α6 or heterodimer αVβ5. Decreased adhesion to fibronectin with integrin α5β1 binding motif, to laminin with integrin α6β1 binding motif and to vitronectin with integrin αVβ5 binding motif, were detected compared with non-inhibition (Fig. 3). These results demonstrate that the plasma membrane of undifferentiated H9 ESC line expresses functional integrin α5β1, α6β1 and αVβ5 heterodimers.

Figure 3. Functional identification of integrin heterodimers expressed in the membrane of undifferentiated human embryonic stem cells (hESCs). The hESCs were incubated in the wells of culture dishes coated with fibronectin, laminin or vitronectin in the absence and presence of anti-integrin α5 (A), anti-integrin α6 (B) or anti-integrin αvβ5 (C) antibodies. The degree of adhesion represents as the optical density of cells plated on poly-l-lysine-coated wells. The hESCs treated with the antibodies showed lower rates of attachment to each extracellular matrix than the non-treated hESCs. All data demonstrate as mean ± SE of three independent experiments. *p < 0.05.

Discussion for Future Perspectives

Integrin expression and signaling in human ESCs has been reported, but the reports have limitation.- It should be redone for the culture of human ESCs in 3D system because there were significant batch effects among human ESC lines. Through this study, we clearly demonstrates expression of integrin α5β1, α6β1 and αVβ5 in undifferentiated human H9 ESC line, which is further identified by functional study using blocking antibodies and ECM-coated culture plate. Characterization of integrin transcription and translation greatly contribute to constructing advanced 3D non-cellular niche for the self-renewal of human ESCs. Sequenced oligopeptides or recombinant proteins with peptide motifs, which binds to characterized integrin heterodimers can be conjugated to 3D PEG-based scaffold.

Based on the results of this study, different culture strategy for the 3D culture can be accepted. For example, it is not necessary to subculture the cells under this in vivo-mimicked, 3D condition. It is more important to maintain ESC characteristics during extended culture period and further characterization is necessary in the ESCs cultured for extended time. Suitable conjugation of ECM analogs to mechanically optimized scaffold increases the potency of synthetic 3D scaffold for maintaining human ESC self-renewal and further confirms the feability of feeder-free, acellular 3D niche for both scientific and clinical purpose.

Based on the results on the expression of integrain subunit, we will employ the motifs for stimulating the integrin signals and the motifs employ have a capacity to stimulate the integrin subunit showing strong expression in the H9 human ESCs. This selection may redo in response to using other human ESC lines because there may be a significant batch effect for the integrin expression. The artificial, acellular niche system using PEG-VS hydrogel and integrin-related peptide mitifs contributes to optimizing extracellular environment for human ESCs. A precise technique to constuct 3D scaffold leads to the development of the techniques for regulating stem cell fate or for bioinjecting into damaged tissue.

Material and Methods

Culture of hESCs

H9 human ESCs were initially cultured on mouse embryonic fibroblast (MEF) feeder layers. The cells were mechanically transferred every 4 to 5 d. The hESCs were cut away from the feeders as small clumps and were subsequently identified by their morphology after collection at further passage. The culture medium consisted of 80% (v/v) DMEM/F12 (Gibco-Invitrogen), 20% (v/v) knockout serum replacement (KSR ; Gibco-Invitrogen), 10 mM nonessential amino acids (Gibco-Invitrogen), 50 μM β-mercaptoethanol (Sigma-Aldrich) and 4 ng/ml basic FGF (Gibco-Invitrogen). Cells were incubated at 37 °C and 95% humidity in a 5% CO2/95% air environment. The culture medium was changed in a daily basis.

Analysis of relative mRNA levels by real-time PCR

The ESCs were transferred to RNAlater® (Ambion) and stored at −75 °C until required. Total mRNA was subsequently extracted using an RNeasy® Mini Kit (Qiagen). The cDNA was synthesized from total RNA using SuperScript™ III First-Strand Synthesis System (Invitrogen). The expression of specific genes was quantified by real-time PCR using a CFX96 Real-Time Detection System (Bio-Rad) using a DyNAmo HS SYBR Green qPCR Kit (Finnzymes Oy). PCR amplification was performed through 40 cycles at 95 °C for 30 sec, 60 °C for 30 sec and 72 °C for 30 sec. Melting curve analysis was performed to check PCR specificity. β-actin expression was measured in each treatment group for standardization, and the expression level of each target mRNA was normalized to that of the β-actin mRNA. Relative mRNA level was calculated as 2−ΔCt, where Ct = the threshold cycle for target amplification and ΔCt = Cttarget gene − Ctinternal reference (β-actin).

FACS analysis

The ESCs were technically collected and suspended by 0.04% (v/v) trypsin-EDTA (Gibco-Invitrogen) treatment. After fixation with 4% (v/v) paraformaldehyde, fixed cells were attached integrin related antibodies for 1 h; integrin α1 (PE conjugated, BD Bioscience), integrin α2 (PE conjugated, BD Bioscience), integrin α4 (PE conjugated, BD bioscience), integrin α5 (PE conjugated, BD Bioscience), integrin α6 (PE conjugated, BD Bioscience), integrin α7 (PE conjugated, R&D Systems), integrin αv (PE conjugated, R&D Systems), integrin β1 (APC conjugated, BD Bioscience), integrin β5 (FITC conjugated, e-Bioscience) and integrin αvβ5 (PE conjugated, Millipore). Prepared samples were measured using FACS Calibur (BD Bioscience) with 20,000 events per sample.

Antibody inhibition assay

Ninety six-well tissue culture plates were coated with 40 μg/ml fibronectin (Sigma-Aldrich), 5 μg/ml laminin (Sigma-Aldrich) and 0.5 μg/ml vitronectin (Sigma-Aldrich) for overnight at 4 °C. Blocking of each well was performed by incubating 10 mg/ml BSA (Sigma-Aldrich) 1 h at 4 °C . Dissociated hESCs resuspended in hESC culture medium containing 2% (v/v) heat-inactivated fetal bovine serum (FBS; HyClone) were plated to each blocked well and incubated in the absence or presence of the anti-integrin α5 (BioLegend), anti-integrin α6 (Millipore) or anti-integrin αvβ5 (Millipore) antibodies for 2 h at 37 °C. Culture dishes were washed with DPBS (Welgene) for removing the non-adherent hESCs from blocked wells. Subsequently, adherent cells were fixed for 10 min in 4% (v/v) paraformaldehyde (Sigma-Aldrich), stained with 0.1% (w/v) crystal violet (Sigma-Aldrich) in 20% (v/v) methanol (Sigma-Aldrich) for 30 min and washed extensively with distilled water. Adherent levels were quantified at 570 nm using a microplate reader (Bio-rad) after adding 50 μl 0.2% (v/v) Triton X-100 (Sigma-Aldrich) in distilled water. BSA (Sigma-Aldrich)- and poly-l-lysine (Sigma-Aldrich)-coated well served as negative and positive control, respectively.

Statistical analysis

All numerical data obtained in each experiment were analyzed statistically using Statistical Analysis System software (SAS Institute). When a significant main effect was detected by analysis of variance (ANOVA) using SAS, the least-squares and Duncan methods were used to compare the different treatments. Level of significant in model and between treatments was determined when p value was less than 0.05.