Advanced Hydrogel for Physiological 3D Colonies of Pluripotent Stem Cells

Abstract

Human induced pluripotent stem cells (hiPSCs) demonstrated great potential in basic research, disease modeling, drug development, cell therapeutics, and regenerative medicine, as various distinct somatic cell types such as hepatocytes can be derived from hiPSCs. However, highly efficient hiPSC to somatic cell differentiation has not yet been achieved because of various challenging problems, one of which is less-optimal culture methods for hiPSC expansion. Conventionally, hiPSCs have been cultured as monolayers on flat surfaces, usually resulting in unstable genetic integrity, reduced pluripotency, and spontaneous differentiation after numerous passages. Recently, three-dimensional (3D) spheroids of hiPSCs have shown potential for somatic cell differentiations. However, these hiPSC spheroids are generated using 2D-cultured cells in either nonadherent U-bottom 96-well plates or agarose microarray molding plates, in which single hiPSCs are forced to aggregate into spheroids. These “aggregation molding” methods are neither typically suited for large-scale hiPSC manufacturing nor for tissue engineering. In addition, the aggregated hiPSC spheroids present limited functions compared to physiologically formed hiPSC 3D colonies. In this chapter, advanced 3D cell culture technologies will be reviewed, and comprehensive discussions and future development will be provided and suggested.

1. Introduction

Human pluripotent stem cells (hPSCs), including human embryonic stem cells (hESCs) and human induced pluripotent stem cells (hiPSCs), are able to self-renew indefinitely in theory and to differentiate into almost all somatic cell types; thus, they have drawn great attention in the research community and hold great potential to improve human health [1]. Unlike hESCs that are obtained from early-stage embryos, hiPSCs can be generated from somatic cells through ectopic expression of defined transcription factors [2] and therefore avoid ethical problems related to the use of hESCs. In addition, the successful generation and specific differentiation of patient-specific hiPSCs provide new approaches for disease modeling, drug screening/toxicity testing as well as personalized cell therapies. However, these applications have been hindered in part by the current less-optimal culture technologies that may reduce or mask some of the concerns related with tumorigenesis, low viability/retention, and uncontrolled in vivo differentiation. Commonly used culture methods cannot completely fulfill the demand for hiPSCs in view of genetic quality, growth performance, stemness, functionality, and differentiation potentiality [3].

Recently, three-dimensional (3D) spheroids of pluripotent stem cells (PSCs), including hiPSCs, have been demonstrated to have great potential for differentiation into various types of organoids, including hepatic cells [4, 5], lung organoids [6], as well as hair-bearing human skin [7]. In most organoid formation studies, the initial spheroids were produced by forced aggregation of single stem cells in suspension/nonadherent plate or agarose microarray molding plates [5, 7, 8]. Despite the advantages observed with 3D differentiation, the starting PSCs were still cultured in 2D, which is not suited for large-scale manufacturing [5, 7]. In this chapter, we will review advanced developments in hiPSC culture methods with emphasis on 3D culture and discuss the benefits and future developments.

2. Culture methods for production of high-quality hiPSC spheroids

2.1 From 2D to 3D cultures

Traditionally, hiPSCs are seeded onto a thin layer of substrate or feeder and grow as colonies of monolayer. Mitotically inactivated mouse embryonic fibroblast (MEF) [9] and Matrigel extracted from mouse sarcoma [10] are the most commonly used feeder layer and substrates for hiPSC culture. For clinical translation, creating a stable, scalable, more defined, and cost-effective culture environment for hiPSC is of high importance. In the past decades, researchers have put tremendous efforts into seeking alternative, well-defined xeno-free matrices, such as laminin [11, 12, 13], E-cadherin [14, 15], fibronectin [16], vitronectin [17], synthetic polymers [18, 19, 20, 21], and synthetic peptides [22, 23, 24, 25]. However, most of these matrices are limited to 2D culture, in which a monolayer of hiPSCs grows on top of matrix-coated flat or bead surfaces. Obviously, these culture systems cannot support large-scale hiPSC production as cell proliferation and expansion requires a large surface area. More importantly, 2D culture is not representing the in vivo physiological environment. The side-to-side cell contact lack of appropriate stem cell niche often leads to unwanted gene expression, resulting in inefficiency of targeted differentiation [26].

Studies on mesenchymal stem cells (MSCs) have shown that by adopting 3D configuration, pluripotency of MSCs was promoted [27, 28]. Similar aggregation techniques were also used before initiating differentiation [5, 7]. However, such suspension methods do not work well for hiPSC culture because of cell agglomeration issues [29, 30, 31]. Other 3D cultures using scaffolds of natural polymers [32, 33] or hydrogel [34, 35] generated in vivo-like conditions for hiPSC culture to avoid cell agglomeration associated with the suspension system. Though the microcarrier bead approach in some of these studies has been considered the “3D” method for large-scale cell manufacturing using stirring tank bioreactors [29, 30, 31, 33], it is still based on 2D cell culture principles. In addition, natural polymer-derived scaffolds generally require a relatively complicated process for cell encapsulation and harvesting, resulting in significant cell loss that would hinder downstream analysis and applications. 3D bioprinting has been explored as a promising method for large-scale production of hiPSC spheroids which allows precise control of spheroid size, but photo-polymerization [36, 37, 38, 39, 40] or temperature control [38, 41, 42] is often required for post-printing gelation. These conditions oftentimes can be harsh for human PSCs [43], leading to low viability and insufficient maintenance of pluripotency.

Hydrogel, on the other hand, is relatively easy to handle. Hydrogels consist of fully synthetic components that not only meet the need to create a well-defined culture environment but also allow the integration of bioactive molecules to promote cellular functions. Lei and Schaffer [35] developed polyethylene glycol (PEG) tailored hydrogel poly(N-isopropylacrylamide)-co-poly(ethylene glycol) (PNIPAAm-PEG or Mebiol Gel™), which is temperature-sensitive and starts gelation at 37°C. This hydrogel was reported to sustain a satisfactory growth rate for hESC and hiPSC lines [35] and maintained pluripotency. However, Mebiol Gel requires high seeding density (1 × 10⁶ cells/ml), and is not easy to use due to its temperature-sensitive nature (it needs to be processed at very low or icy temperatures [35, 44, 45]). In addition, cell viability in Mebiol Gel was less than 20% [46], which is lower than required for various applications. Therefore, a practical and well-defined 3D matrix system that is proficient at maintaining high-quality hiPSCs for long-term culture and downstream applications remains to be developed.

2.2 Synthetic peptide hydrogel

A tri-block amphiphilic peptide-based hydrogel termed as h9e was discovered by Huang et al. [47] and Sun and Huang [48]. H9e is rationally designed from a group of selective amino acids that can self-assemble into nanofibers and then transform into a fast sol–gel reversible hydrogel through shearing force, such as pipetting or syringing, under neutral pH at room or body temperature [48, 49]. H9e’s peptide sequence originated from two functional native proteins of human muscle [50] and the β-spiral motif of the spider flagelliform silk protein; therefore, it is highly compatible with biological systems [51]. This peptide is also reconcilable with various cell culture media, such as DMEM, MEM, RPMI, and L-15, as well as common hiPSC culture medium mTeSR and Essential 8 (E8). Through modification of the backbone structure, a variety of h9e can be created with the desirable hydrogel properties to meet specific requirements for 3D cell culture or in vivo delivery. For instance, PGmatrix system, a commercial product derived from h9e by PepGel LLC (Manhattan, KS), has been used for 3D cultures of various cancer cells [52, 53, 54, 55] and hiPSC [46], in vivo delivery of drugs, antigens, viruses [47, 56, 57] as well as human MSCs [58] with PGmatrix were reported.

2.2.1 Mechanical properties of peptide hydrogel

PGmatrix is composed of entangled nanofibers forming a porous scaffold as shown in Atomic force microscopy (AFM) images (Figure 1). Single nanofiber with a 20 nm in diameter and nanofiber clusters with 100–500 nm in diameter were identified. The pore size was up to 2.5 μm. This peptide hydrogel is compatible with mTeSR1, the commonly used culture medium for hiPSCs, and formed a self-supporting hydrogel in a similar manner as reported by Liang et al. [55]. The self-assembling nature of PGmatrix was observed by measuring gel formation as a function of time. Within a few seconds, gel strength reached 100 Pa at the concentrations of 0.5% and 1.0% peptide; shear-thinning and self-recovery properties were measured at peptide concentrations of 0.2%, 0.5%, and 1.0% (Figure 2). After 1 minute of shear-thinning treatment, PGmatrix gel transformed into liquid-like status with about 0.1–0.5 Pa of elastic moduli. Once the shear-thinning force was removed, it quickly recovered to gel status. After 1 minute, the gel recovered to 67%, 70%, and 80% of its original strength in the samples with 0.2%, 0.5%, and 1.0% concentrations of peptide, respectively. After 10 minutes, all three samples restored up to 95% of the original gel strength (Figure 2). These unique properties suggest that PGmatrix scaffold can be easily manipulated manually or automatically by high throughput robot or bioprinting for cell encapsulation, expansion, and mechanical isolation.

Figure 1.

AFM images of the PGmatrix nanostructure at low (left) and high (right) resolutions.

Figure 2.

Storage moduli of PGmatrix were directly proportional to the concentrations of peptide nanofibers in both shear-thinning and sol–gel recovery tests.

2.2.2 Peptide hydrogel for long term hiPSC culture

hiPSCs derived from foreskin fibroblasts were cultured in 0.5% PGmatrix by Li et al. [46] in parallel with 2D culture in Matrigel-coated plates. hiPSCs grown in 3D PGmatrix showed significantly higher fold expansion (p = 0.014) and cell viability (p < 0.0001) (Figure 3A) compared to those on 2D Matrigel (Figure 3B). The variance of fold expansion across multiple passages (P13–P25) of hiPSC was significantly smaller in 3D (15.70 ± 3.70) than in 2D (10.55 ± 7.05) (p < 0.05). Another hiPSC line derived from CD44+ somatic cells was cultured in 3D PGmatrix as well as in 2D for multiple passages. Similarly, less variance in fold expansion and viability was observed with 3D culture [46].

Figure 3.

Growth performance of hiPSCs in 3D PGmatrix (A) and in 2D on Matrigel (B). Data are shown as means ± SDs.

After long-term culture in 3D PGmatrix, hiPSCs showed similar expression of Oct4, Nanog, Sox2, and SSEA4, which are comparable to those in 2D-cultured cells, however, differentiation marker gene AFP and Brachyury were significantly less expressed in 3D-cultured hiPSCs (Figure 4) [46]. The pluripotency of hiPSCs cultured in 3D PGmatrix was also verified with teratoma formation assay. In addition, these hiPSCs retained normal karyotype after long-term maintenance in 3D PGmatrix [46].

Figure 4.

hiPSCs showed upregulation of pluripotency-related genes in 3D PGmatrix. 2D results were averaged from two different passages (P10 and P15), while 3D results were averaged from four different passages (P7, P10, P15, and P25). Data are shown as means ± SDs. *p < 0.05.

However, it is interesting to note that compared to the expression levels of SSEA4 and TRA-1-81 in pooled hPSC lines on 2D culture as reported by the International Stem Cell Initiative [59], TRA-1-81 expression (~7%) was significantly lower in two hiPSC lines cultured in 3D PGmatrix hydrogel. Although SSEA4 and TRA-1-81 have been commonly used as hiPSC markers, the TRA-1-81 expression may not be crucial in the maintenance of hiPSC pluripotency. It has been reported that naïve-state hPSCs did not express SSEA4 in both TRA-1-81 negative and positive fractions [60, 61]. Even though both SSEA4 and TRA-1-81 antigens are believed to be involved in cell adhesion, the mechanisms underlying hiPSC maintenance in 3D PGmatrix might be different from that of the conventional 2D culture. It is possible that high expression of SSEA4 and low expression of TRA-1-81 observed in the present 3D culture is due to direct physical or biological interactions between h9e peptide and hiPSCs. Nevertheless, the findings from this study suggest that some of the PSC markers established in 2D culture may need to be revised for the characterization of 3D-cultured hiPSCs.

On the other hand, differential expressions of certain genes were identified in 3D-cultured hiPSCs. Significant upregulation of UTF1 and hTERT, but downregulation of REX1, FGF4, and GDF3 were observed for both hiPSC lines in 3D compared to 2D culture (Figure 4). Being a PSC gene, UTF1 is closely associated with stem cell pluripotency. Researchers found that UTF1 may play an important role in chromatin formation of embryonic stem cell (ESC) [62]. Another critical factor controlling the proliferation of PSCs is the integrity of telomeres, which is regulated by hTERT transcriptor [63]. Overexpression of hTERT in hMSCs increased differentiation potential and decreased spontaneous differentiation [64]. Therefore, high expression levels of UTF1 and hTERT in the present 3D hiPSCs may imply high-quality hiPSCs in the aspects of genetic integrity, pluripotency, and proliferation compared to those in 2D systems. REX1 (also known as Zfp42) is a zinc-finger encoding gene expressed exclusively in early embryos and has been widely used as a PSC marker [65]. Early studies demonstrated that knockout of REX1 (REX1-/-) in ESCs does not affect cell proliferation and pluripotency [66, 67]. A later study discovered that REX1 regulates human stem cell pluripotency by promoting mitochondrial fission, which keeps mitochondria in an immature state and stem cells in a highly glycolytic state [68]. This function of REX1 is critical to protecting hiPSCs in 2D because a high oxygen level in 2D culture would trigger mitochondrial oxidative phosphorylation. However, in the 3D PGmatrix system, oxygen level might be limited by hydrogel as a “diffusion barrier”; therefore, upregulation of REX1 would be less critical for cells in 3D culture.

2.3 Comparison of hydrogels for 3D hiPSC culture

Since the rapid growth and high viability of hiPSCs cultured in 3D PGmatrix might be simply resulted from a 3D culture environment, 3D culture in PGmatrix was compared with PNIPAAm-PEG, Mebiol Gel [35] in parallel using the same hiPSCs using either mTeSR1 or E8 medium [46]. Under optimal conditions (PG-mTeSR1 vs. Mebiol-E8), hiPSCs formed a more uniform spherical morphology than those in Mebiol gel. Spheroids were larger in 3D Mebiol gel, but cells on the edge of the large spheroids seemed to be dying (Figure 5). Pluripotency gene expression compared by RT-qPCR showed that spheroids formed in the 3D Mebiol gel had lower pluripotency than those in the 3D PGmatrix [46]. These findings suggested that hiPSC maintenance performance is not only linked to 3D cell conformation and hypoxic condition provided by hydrogel. By comparing gel strength kinetics, gel degradability may be a contributing factor to hiPSC growth [46]. It has been proved that matrix degradation is crucial for the maintenance of neural progenitor cell stemness [69], the mechanosensing genes yes-associated protein 1 (YAP) and transcriptional coactivator with PDZ-binding motif (TAZ) [69] are at the center of this signaling cascade, which was also confirmed by Li et al. [46]. The maintenance of hiPSC pluripotency may be related to PGmatrix’s degradability that allows encapsulated hiPSCs to modify their environment. Soluble factors such as insulin, bFGF, TGFβ, and nodal from medium supplements would bind to cell surface receptors such as GPCRs (G protein-coupled receptors) and RTKs (receptor tyrosine kinases) to activate essential pathways, including phosphoinositide-3-kinase–protein kinase B/Akt (PI3K/Akt) for survival [70, 71]. During this adaptation, hiPSCs may initiate their secretion of extracellular matrix (ECM) proteins and some proteases and start remodeling the surrounding environment through matrix degradation and modification. Secreted ECM proteins such as laminin and vitronectin would in turn bind to integrin and trigger various downstream signaling pathways to promote proliferation and pluripotency maintenance [72, 73, 74]. The microenvironment of ECM-protein-modified-PGmatrix would facilitate cell migration, leading to increased cell–cell contact via E-cadherin [75, 76, 77]. These binding events are closely linked to cell cytoskeleton, affecting actin dynamics and relaying the signals through Hippo pathway. Mechanical signals transduced through the signaling cascade, including LATS 1/2 would cause upregulation of the mechanosensitive YAP and TAZ proteins, which then relocate into the nucleus to affect gene expression related to pluripotency maintenance (Figure 6) [46, 78, 79, 80, 81, 82].

Figure 5.

Morphology of hiPSCs grown in PG-mTeSR1 and Mebiol-E8 on days 0, 2, and 4 after encapsulation. Scale bar, 50 μm.

Figure 6.

Proposed mechanism for hiPSC growth and maintenance of pluripotency in 3D PGmatrix hydrogel. (Reprinted from Li et al. [46] with permission.)

2.4 Develop peptide hydrogel for hiPSC bioprinting

To meet the need for 3D bioprinting, PGmatrix peptide hydrogel was modified into PGmatrix-M bioink that self-heals to a gel state after printing without any crosslinking aids from UV light or chemicals. In addition, PGmatrix-M supports hiPSC proliferation and maintenance after printing with viability above 95% [46]. It has been demonstrated that after 3D bioprinting, hiPSC aggregated can be differentiated into various tissues, including cartilage [83], hepatocyte-like cells [84], and neural tissues [85]. Compared to bioinks used in these studies, PGmatrix-M offered a gentler environment that allows long-term hiPSC maintenance. This raised the possibility of culture patterned hiPSC spheroids after printing and perform more complex differentiation processes, which would greatly benefit tissue engineering to produce functional organoids.

3. Conclusion

We have reviewed the advanced development in 3D hiPSC culture systems that have replaced the initial complex feeder layer or protein mixture Matrigel. Among the 3D technologies, the innovative peptide hydrogel PGmatrix outperformed its predecessors in terms of long-term hiPSC maintenance and the ability to be easily handled under ambient conditions. It is believed that usage of this 3D platform in the hiPSC field will not only promote the production of high-quality hiPSCs, which could lead to improvement of differentiation efficiency in generations of various types of somatic cells, including hepatocytes, but also provides a new tool to manufacture hiPSC at an industrial scale for downstream applications. In addition, the classical stem cell markers, such as TRA-1-81, should be reevaluated regarding their sensitivity in characterizing hiPSCs in 3D culture.

Conflict of interest

The authors declare no conflict of interest.

Appendices and nomenclature

AFP	alpha-fetoprotein
AFM	atomic force microscopy
bFGF	basic fibroblast growth factor
ECM	extracellular matrix
GPCRs	G protein-coupled receptors
hESCs	human embryonic stem cells
hiPSCs	human induced pluripotent stem cells
hPSCs	human pluripotent stem cells
MSCs	mesenchymal stem cells
MEF	mouse embryonic fibroblast
PI3K/Akt	phosphoinositide-3-kinase–protein kinase B/Akt
PEG	polyethylene glycol
PNIPAAm-PEG	poly(N-isopropylacrylamide)-co-poly(ethylene glycol)
RTKs	receptor tyrosine kinases
RT-qPCR	reverse transcription-quantitative polymerase chain reaction
ROCK	rho-associated, coiled-coil-containing protein kinase
3D	three-dimension
TGFβ	transforming growth factor-beta
TAZ	transcriptional coactivator with PDZ-binding motif
YAP	yes-associated protein 1