1. Introduction

Human Pluripotent stem cells (hPSCs), including hESCs that first isolated by Thomson et al. [1] in 1998 and hiPSCs established in 2006–2007 by Yamanaka’s group [2,3], retain the core characteristics of self‑renewal and pluripotency, the capacity to differentiate into all three germ layers. So that, these unique properties have established hPSCs as a versatile platform for studying disease modeling, drug screening and regenerative medicine.

Early hPSC differentiation strategies, such as embryoid body formation and feeder-layer co-culture with mitotically inactivated mouse embryonic fibroblasts (MEFs), provided proof of concept for in vitro lineage specification by supplying extracellular matrix components and soluble inductive cues [1,4]. At that time, MEF feeders were essential for maintaining hESC pluripotency due to the absence of fully defined culture systems. However, the development of feeder-free, chemically defined media—such as mTeSR1 [5] and E8 medium [6]—combined with human-derived extracellular matrix coatings (e.g., vitronectin, laminin-521) has eliminated the need for mouse-derived feeder cells, enabling xeno-free culture conditions suitable for clinical-grade cell manufacturing. These advances, together with the identification of lineage-inducing factors—key transcription factors (e.g., SOX17, GATA4), growth factors, and small molecules like CHIR99021 —have greatly improved differentiation efficiency and reproducibility.

Because hPSCs have the inherent ability for unlimited self-renewal, theoretically they can be infinitely expanded and continuously differentiate into the target cells. Therefore, various cell types derived from hPSC (human pluripotent stem cells) can be produced on a large scale. Traditional drug screening often relies on animal models or limited human samples. However, hPSCs can continuously provide human-derived functional cells (such as neurons and cardiomyocytes), which are closer to the human physiological state. Such characteristics make them suitable for use in high-throughput drug screening, and hPSCs (especially iPSCs from patients themselves) can also be used for the development of personalized cell therapy [7].

Nevertheless, there are still challenges, including suboptimal differentiation efficiency, cellular heterogeneity, and incomplete functional maturation in vivo. In addition, for hESCs, unresolved ethical concerns surrounding embryo use continue to represent a major barrier to their clinical application.

Here, this paper overviews the most recent progress in understanding the underlying mechanisms of hPSC differentiation with emphasis on gene-editing technologies for directed differentiation, using small molecules as modulators, and 3D culture systems.

2. Gene editing technologies in hPSC differentiation

CRISPR/Cas9 technology has transformed the study of hPSCs by providing a versatile tool for precise and efficient genome editing. González et al. first established an inducible CRISPR (iCRISPR) platform in hESCs and hiPSCs, demonstrating robust and multiplexable gene editing [8]. Then inducible CRISPR interference and activation systems (CRISPRi/a) were developed, which enabled reversible and fine-tuned gene regulation in hPSCs [9]. Mandegar et al. (2016) proved that CRISPRi targeting OCT4 efficiently reduced its expression by more than 90%, making to loss of pluripotency and spontaneous differentiation, while CRISPRa targeting NANOG upregulated its transcription and then enhancing stem cell identity. Importantly, both effects were reversible upon doxycycline withdrawal, highlighting the potential of CRISPRi/a for finding gene function in hPSCs. As added in Figure 1 (A&B) and Figure 2, Mandegar et al. established doxycycline-inducible CRISPRi/a systems in hPSCs, presenting reversible regulation of Cas9 expression and functional control of pluripotency genes.

Subsequently, large-scale CRISPR-based functional genomic screens, often integrated with single-cell transcriptomics, have allowed systematic dissection of lineage regulators and stress response pathways. For example, Adamson et al. developed a multiplexed single-cell CRISPR screening platform that enabled high-resolution mapping of the unfolded protein response (UPR), providing a paradigm for how CRISPR perturbations can be systematically linked to transcriptional states at single-cell resolution [10].

Together, these methodological advances illustrate a trajectory from basic gene targeting towards scalable and high-resolution functional genomics, laying the foundation for rational optimization of hPSC differentiation protocols.

3. Small-molecule regulation of hPSC differentiation

Small molecules offer a powerful, cost-effective, and scalable means to direct hPSC differentiation with high reproducibility. Unlike growth factors, which suffer from high cost and batch variability, small molecules precisely modulate intracellular signaling pathways. For example, CHIR99021, a GSK3β inhibitor, robustly activates Wnt/β-catenin signaling to promote mesoderm induction, whereas SB431542 blocks Activin/Nodal signaling to prevent endodermal or ectodermal diversion.

A typical example of small-molecule regulation is the cardiac differentiation protocol reported by Lian et al., which modulates Wnt signaling in a stepwise manner using the GSK3β inhibitor CHIR99021 followed by the Wnt inhibitor IWP2 under chemically defined conditions [11]. After 24 h of CHIR99021 treatment, ~98% of hESCs became Brachyury-positive, indicating highly efficient mesodermal induction (Figure 3A). By day 6, cells expressed the cardiac progenitor marker Isl1, confirming differentiation toward the cardiac mesoderm (Figure 3B).

The overall time course of differentiation is summarized in Figure 4: hPSCs (Oct4+) transition into mesoderm (Brachyury+), then into cardiac mesoderm (Isl1+), and subsequently cardiac progenitors (Nkx2.5+/Isl1+), finally leading to >80% cTnT+ cardiomyocytes in day 15.

The study shows that, structural analysis of day 20 hPSC-derived cardiomyocytes revealed well-organized sarcomeric evidenced by immunostaining for α-actinin and MLC2a (Figure 5), proofing not only the efficiency but also the functional maturation of the differentiated cells.

Overall, this example highlights that staged small-molecule modulation of Wnt signaling provides a simple, reproducible, and growth factor–free strategy for directing hPSC differentiation into cardiomyocytes.

Similarly, Ma et al. reviewed that CHIR99021 combined with SB431542 significantly enhances lineage-specific outcomes across multiple protocols [12]. To sum up, small-molecule approaches enable more defined, reproducible, and clinically adaptable differentiation strategies.

4. 3D systems for enhancing hPSC differentiation

Three-dimensional (3D) culture systems better recapitulate the in vivo microenvironment compared to traditional two-dimensional monolayer cultures. By providing spatial organization, extracellular matrix interactions, and physiological gradients, 3D models improve differentiation efficiency and functional maturation.

For example, Lancaster et al. established cerebral organoids from hPSCs that mimic human cortical development and enable disease modeling of microcephaly [13].

They developed a cerebral organoid culture system by embedding hPSCs in Matrigel droplets and maintaining them in spinning bioreactors, which provided sufficient nutrient and oxygen exchange to support long-term growth. Brain tissues developed by this way showed a very high efficiency, which needs only 8-10 days to make the appearance of neural identity and form defined brain regions in only 20-30 days(Figure 6) .Within the cortical regions, radial glia, neural progenitors, and layered cortical neurons were observed, recapitulating key features of early human brain development. Importantly, when patient-derived iPSCs carrying CDK5RAP2 mutations were used, the resulting organoids were significantly smaller because of damaged progenitor proliferation, thus modeling the pathological features of microcephaly. This work established cerebral organoids as a powerful 3D platform for studying human neurodevelopment and neurodevelopmental disorders.

Similarly, Zhang et al. developed cardiac organoids with tissue-level organization using hPSC-derived cardiomyocytes and endothelial cells on 3D bioprinted scaffolds [14]. Building on these advances, Ronaldson-Bouchard et al. further demonstrated that applying mechanical and electrical stimulation to hPSC-derived cardiac tissues promotes advanced maturation, yielding structural, electrophysiological, and metabolic properties closer to those of adult human myocardium [15]. These 3D platforms provide not only mechanistic insights into human development but also translational models for drug testing and regenerative medicine.

5. Discussion

In recent years, gene editing technology, small molecule regulatory strategies, and three-dimensional culture system technologies have each provided different but complementary perspectives. The CRISPR-based gene editing platform enables researchers to truly pose causal questions: What role does a certain transcription factor, epigenetic regulatory factor, or signaling pathway play in the differentiation process, rather than merely making correlation observations? The small molecule regulation methods have fully demonstrated that precise regulation of developmental signals not only can replace expensive exogenous growth factors but also can improve the stability and reproducibility of experiments. The three-dimensional model also indicates that the tissues derived from hPSCs have the potential for self-organization and can form the same spatial structure as in the body, to a certain extent, they can also reproduce the key features of human development and even have certain advantages over traditional animal models in disease modeling. However, there are still some limitations. The differentiated cells generated by hPSCs are usually closer to the fetal state rather than mature adult cells, thus limiting their direct application in regenerative medicine. Although CRISPR technology has great potential, it still has safety issues, such as off-target effects and genomic stability, especially when considering clinical translation, more caution is needed. Although organoids have significantly provided a window for studying human development, they still lack the support of other systems. Therefore, although the hPSC differentiation research has made significant progress, it is still quite far from "completely solving" the problem.

In the future, the combination of these three cutting-edge technologies will not only facilitate a deeper understanding of human also lay the foundation for the clinical application prospects of pluripotent stem cells.

6. Conclusion

Human pluripotent stem cells (hPSCs), which include both human embryonic stem cells (hESCs) and human induced pluripotent stem cells (hiPSCs), have the capacity for infinite self-renewal and differentiation into virtually any tissue of the body, defining an unparalleled platform for regenerative medicine, disease modeling, as well as high-throughput screening.

In the last few years, significant progress has been reached in unraveling the molecular principles of hPSC differentiation and in optimizing culture systems to increase efficiency and standardization. These include recent developments in CRISPR/Cas9-mediated genome engineering, small-molecule–based lineage control, and three-dimensional differentiation platforms to direct hPSC fate at higher resolution and with increased physiological relevance.

However, there are challenges remaining in need of resolution, such as lack of full functional maturation of derived cells, cellular contamination by hESCs, noted genotype variability, and issues surrounding the ethics of using hESCs. To this end, the future will have to overcome these limitations by integrating multi-omics approaches, genome editing, as well as recent bioengineering innovations.

In conclusion, improvement in hPSC differentiation, such as chemically defined media, will be essential for translating stem cell-based therapies to the clinic, opening a new era of personalized regenerative medicine.