Engineering pluripotent stem cells with synthetic biology for regenerative medicine

Immune rejection

One of the major safety concerns that need to be addressed prior to widespread clinical use of PSCs and their derivatives is immunogenicity [34]. Immune rejection is a major obstacle for allogeneic cell transplantation, mainly due to human leukocyte antigen (HLA) mismatch between the transplanted cells and the recipient. Besides HLA mismatch, other reasons include immune antigens presentation by PSCs due to prolonged in vitro culture, atypical antigens formation due to genomic mutations, incomplete reprogramming of somatic cells during iPSC reprogramming, and immune immaturity of PSC-derived cells [34]. Autologous, patient-specific iPSC-derived cells may overcome these challenges. However, autologous iPSCs are too costly given the necessary quality control and too slow to meet acute needs. An alternative strategy is to use HLA-matched, quality-controlled iPSC banks that are being developed in Japan [34]. Such banks could provide standardized sources of cells for clinical applications, overcoming the need for individualized autologous cell lines and addressing cost and scalability concerns. An additional approach is HLA cloaking, which involves inactivating HLA genes and/or overexpressing immunosuppressive factors in allogeneic stem cells [35, 36]. This approach can reduce the immunogenicity of allogeneic stem cells, making them more suitable as “off-the-shelf” cell therapy products for large-scale clinical applications. Only a very small number of hPSC lines or even one single line could be required to serve as a universal and standardized source of cells covering the entire world population. Such universal hPSCs would be allogeneic, more homogeneous, cheaper, and safer, making them better candidates for engineering and serving as off-the-shelf cell therapy products for large-scale clinical applications.

Tumorigenicity

Self-renewal ability is one of the important properties of hPSCs, but also a double-edged sword since its potential tumorigenesis if the cells continue to proliferate indefinitely after transplantation [34]. Besides, progenitor cells used for cell therapy also have been reported to exhibit the tumorigenic potential after transplantation. In addition, tumorigenesis can be caused by several other reasons, including cells with unexpected high proliferation capacity, genomic abnormalities due to long-term culture and gene-editing manipulations. Recently, a systematic study pointed out that over 20 % of both hPSCs and their in vitro derivatives harbored cancer-related mutation (s). Particularly, 64 % of these samples contained mutations in the TP53 gene, which is a common tumor suppressor gene and its dysregulation is highly correlated to tumorigenesis and cancer progression [37]. These results highlight the critical importance of vigilantly monitoring for cancer-related mutations in hPSCs, especially when they are being considered for clinical use. Since the potential tumorigenicity of hPSCs and their derivatives still cannot be completely ablated in vitro by genetic engineering, dynamically monitoring and eliminating the cells with tumorigenic transition is a practical approach to ensure the safety of the cell therapy product. For example, stem cells can be engineered with inducible suicide or elimination switches, such as fail-safe systems based on suicide genes Caspase9 and HSVtk [38, 39] and miRNA switches [40].

Robust and homogeneous differentiation of specific cell types that can survival, proliferate and function maturely

The differentiation of PSCs into desired cell types is a primary step in the development of cell-based therapies. However, there are significant challenges associated with the low efficiency and reproducibility of this process, resulting in generating cells with low purity, immature function and limited reproducibility. Additionally, for cell replacement therapy, transplanted cells must possess the sufficient survival and proliferation capacity, followed by homing and localizing to the therapeutic site and executing the therapeutic functions. Importantly, donor cells are hard to communicate and integrate with recipient tissues to initiate pre-designed functions. Taken the clinical trials of PSC-derived RPE cells to treat age-related macular degeneration (AMD) as an example, the transplanted RPE cells failed to precisely home and localize to the therapeutic site. Another example is the PSCs-derived dopaminergic progenitor cells-based PD therapy, the major challenges are durability and survival of cells, synaptic integration and functional restoration after transplantation [41].

Inefficiency of scale-up and quality control for cell manufacture

Similar to the manufacture of other biologics, that of hPSCs and their derivatives should be strictly subjected to the instructions of good manufacturing practice (GMP), which provides the standards for safety, efficacy and feasibility of therapeutic cells for large scale production [41]. Heterogeneity of hPSCs is reflected in the clonal variability of the cell lines, difference between cells within each clonal line during prolonged in vitro culture, as well as the continuous fluctuation of gene and protein expression levels at a frequency ranging from hours to days [34]. The hPSCs might undergo genetic mutations during long-term culture [34]. In addition, the differentiation of hPSCs in vitro can generate heterogenic subpopulation [34]. Therefore, development of effective and less labor-intensive approaches to regulate the purity and genetic background of hPSCs is important for product safety. Recently, the cost-effective scale-up in bioreactors has been used to overcome cumbersome labor-intensive production [42].

Low efficiency and reproducibility of differentiation

The differentiation of PSCs into desired cell types is a critical step in the development of PSC-based therapies. The predominant way to differentiate PSCs into various cell types is stepwise addition of specific growth factors, small molecules, and signaling inhibitors to mimic the corresponding developmental extracellular environment. However, there are significant challenges associated with the low efficiency and reproducibility of this process. Low efficiency can lead to the formation of cells that are functionally immature and not therapeutically relevant or that may even have tumorigenic potential, while lack of reproducibility can result in cell products that are variable in their potency and functionality. For example, many studies have reported that hESC-derived cardiomyocytes express marker TFs, show a cardiomyocyte-like phenotype, while display immature electrical phenotype [41]. Assurance of efficacy relies on a robust and reliable set of PSC differentiation protocols. Efforts to address these issues include the development of new culture conditions and genetic manipulation strategies that aim to improve cell differentiation and to ensure that the resulting cells are mature. Many groups have reported that modification of the timing, concentration and types of the growth factors, small molecules, signaling inhibitors can promote the efficiency and reproducibility of differentiation; while some other groups have reported that over-expression of master TFs can enhance cell differentiation [43], [44], [45], [46], [47]. In addition, efforts have been made to develop new culture platforms that allow for greater control over cell differentiation, including 3D bioreactors and microfluidics devices that mimic physiological environment [42]. Finally, a more effective solution is obtaining in-depth insight into the genetic regulatory networks governing stem cell differentiation in vivo and engineering them during in vitro differentiation. Analysis approaches combining gene expression data, proteomic pathway data, and cell signaling models have been created to explore key state transitions during differentiation [48].

Disordered organ-specific morphological architecture

Although organoids can display organized structure pattern compared with 2D monolayer cells, recapitulation of organ-specific spatiotemporal architecture still cannot be realized. Lack of necessary lineage cells like stroma cells and vascular endothelial cells is one reason for disordered structure. Another important reason is the absence of in vitro morphogen environment which plays a key role in developmental tissue organization in vivo. Providing morphogen gradient in vitro is promising to improve the structure of organoids, by either cell engineering, 3D-bioprinting, or other novel techniques.

Immature function

The predominant advantage of organoids is the improved tissue-like function filling the gap between 2D monolayer cell with limited function and the animal models. However, to date, majority of hPSC-derived organoids exhibit fetal cell-type composition and function analyzed by multi-omics, such as the liver organoids with fetal-like hepatocyte activity including cytochrome P459, vitamin A storage, lipid accumulation and inflammation response [53]; the cerebral cortical organoids resembling 19–24 post-conception-week prenatal brain [54]. Limited studies have reported to establish the PSC-derived organoids with neonatal function. In 2021, the establishment of hypothalamic arcuate organoids was reported with high similarity to neonatal human hypothalamus in cell type diversity and molecular signature, but whether possess the neonatal-level function was not demonstrated [55]. Optimization of differentiation protocols, integration of other lineages, re-construction of microenvironment might be the underlying strategy to overcome these challenges.

Limited size and lifespan

The limited lifespan and size of organoids are two related but different challenges in the field of organoid research. The limited lifespan and size are mainly caused by the limited nutrient supply and waste removal that occur as the organoid grows over time [56]. As organoids increase in size, the efficiency of nutrient supply and waste removal through diffusion decreases, which can give rise to limited growth, cell death and organoid necrosis. To address these challenges, researchers have developed various strategies such as using larger bioreactors [42] or using a perfusion system to improve nutrient access and waste removal [57]; improving vascularization to further spread nutrients through capillaries.

To realize organ or tissue replacement by PSC-derived organoids, development of innovative technologies, as well as combination with other disciplines are necessary to overcome some of the unsatisfactory outcomes of current organoids and further address the barriers.

Apart from the organoids, another type of multicellular system, called “synthetic embryos” including human blastoids has been established. Successful establishment of human blastoids were reported in 2021 by several groups, respectively [17], [18], [19], [20, 22]. Human blastoids were created via aggregation of ESCs, self-organization of naïve ESCs [17, 18], or even reprogramming of fibroblasts [20], respectively. Besides, another group generated human naïve PSC-derived blastoids which mimics the sequential lineage specification (trophectoderm, epiblast and primitive endoderm) of blastocyst development, and directionally attaches to endometrial cells with the polar trophectoderm to model implantation [58]. These synthetic embryos provide important multicellular models for developmental modeling, disease modeling, drug screening and organ regeneration in vitro.

To summarize, a major barrier to the clinical translation of PSC bioengineering efforts has been our achievement of a holistic understanding of the gene regulatory networks (GRNs) and biological processes governing cell behaviors inability to predictably and reproducibly control individual stem cell behaviors and their self-organization into multicellular systems. Stem cell engineering by synthetic biology can push the solutions.

Precise control of cell fate decision

One of the prominent advantages of hPSCs for cell therapy is the differentiation potency into various cell types, especially the cells which are hard to obtain from human tissues or possess limited proliferative capacity in vitro. In view of this, hPSCs have been regarded as a vital source to construct the pools of various types of cells by robust differentiation for cell therapy. However, based on current differentiation strategy using cocktail of small molecules and growth factor inhibitors, majority of hPSC-derived cells cannot be applied in clinical trials due to the immature function, low efficiency and high heterogeneity.

Direct differentiation of hPSCs into certain cell types or further organoids using forward programming approaches, including forced expression of TFs, provides a potential alternative. For example, thyroid follicular cells and transplantable organoids were differentiated from mouse and human PSCs through transient overexpression of NK2 homeobox 1 (NKX2-1) and paired box gene 8 (PAX8) [43, 44]. Besides, functional thyroid progenitors were robustly differentiated from mouse PSC-derived anterior foregut endoderm via transient and developmental stage-specific overexpression of NKX2-1 [47]. In addition, production of hPSC-derived megakaryocytes by overexpression of GATA1 (GATA binding protein 1), FLI1 (Fli-1 proto-oncogene, ETS transcription factor) and TAL1 (TAL bHLH transcription factor 1, erythroid differentiation factor) were reported as well [46]. Another group reported robust differentiating hPSCs into neurons, functional skeletal myocytes and oligodendrocytes [45].

However, such overexpression methods cannot precisely and robustly regulate cell fates because they fail in spatiotemporal controlling of gene expression and precisely controlling dosage of gene products. Precise encoding cell fate decision by synthetic lineage-controlling circuits holds great potential to obtain targeted cells with high efficiency and purity, facilitating their clinical application and scale-up production (Figure 4).

Figure 4:

Control of PSC differentiation using complex synthetic lineage-control circuits. Schematic of synthetic genetic regulatory network programming differential expression dynamics of master transcription factors during hPSC differentiation. The three circuits provide programmable, mutually exclusive expression switches for gene A (ON-OFF-OFF), gene B (OFF-ON); gene C (OFF-ON-OFF), gene D (OFF-ON-OFF); and gene E (OFF-ON) expression in a temporal manner. These synthetic genetic circuits synergistically direct the differentiation of hPSCs towards specific cell types step by step.

A proof-of-concept work pointed out that synthetic genetic circuit would be a promising strategy to obtain hPSC-derived cells by directly controlling expression of master genes during development [65]. The design of lineage-control circuit was based on the previous studies of mouse pancreatic development, which demonstrated that the temporal expression pattern of neurogenin 3 (NGN3), pancreatic and duodenal homeobox 1 (PDX1) and V-maf musculoaponeurotic fibrosarcoma oncogene homolog A (MAFA) play a crucial role in the differentiation and maturation of pancreatic β cells. Thus, researchers designed a synthetic lineage-control network capable of mimicking the expression patterns of NGN3 (OFF-ON-OFF), PDX1 (ON-OFF-ON), and MAFA (OFF-ON) during in vivo pancreatic development by utilizing a vanillic acid dose-dependent signaling cascade and a gene switch. In hiPSC-derived pancreatic progenitor cells, this network enables correctly temporal differential expression of these three TFs and ultimately differentiates cells into glucose-sensitive insulin-secreting β-like cells. Notably, compared with cocktail-induced differentiation, these pancreatic β-like cells exhibited more similar dynamics of glucose-stimulated insulin release to that of human pancreatic islets, indicating the great potential of using synthetic gene circuits to establish the pool of functional cells for cell therapy (Figure 4).

Efficient purification of cell populations during PSC differentiation

The resulting heterogeneous population is a major problem of PSC differentiation. Thus, detection and purification of specific cell types for therapy is critical. Synthetic cell classifiers, usually based on microRNA (miRNA) switches, were created to detect and purify cells of interest. In 2015, Miki reported the establishment of synthetic miRNA switches to isolate desired cell type during hPSC differentiation [66]. Various miRNA switches were created, including miR-1,208a and 499a-5p for cardiomyocytes, miR-126 for endothelial cells, miR-122-5p for hepatocytes, and miR-375 for insulin-producing cells [66]. In detail, the miR-1 switch contains a synthetic mRNA encoding a fluorescent protein (FP) tagged with target site of miR-1, and desired cells expressing endogenous miR-1 will not express FP. And the miR-1-Bim switch, containing the apoptosis inducer Bim, will induce selective apoptosis in cells without miR-1 expression, namely non-target cells.

Enforcing enhanced or artificial therapeutic functions

Beyond achieving physiological functions, therapeutic function can be further enhanced, such as targeting and killing tumor cells. Predictable and defined levels of therapeutic functions must be achieved either by rewiring existing networks or engineering new ones. Synthetic biology is poised to enhance cell-based therapies by offering precise control over the context, intensity, and timing of therapeutic intervention. The therapeutic outputs of cells should sense inputs such as small molecules and disease biomarkers. After the signal processing, the dosage, timing, and localization of therapeutic gene expression or other therapeutic activities (outputs) should be tightly controlled.

A big milestone in this field was the FDA approval of Tisagenlecleucel, a CAR T cell product, in 2017 to treat refractory pre-B cell acute lymphoblastic leukemia and diffuse large B cell lymphoma. Specific immune cells including T cells and natural killer (NK) cells are key players of antitumor immunity, and the synthetic receptor CAR [54] can furthermore enhance their anti-tumor function. Such therapy consists of immune T cells taken from the patient, which are then engineered to express CAR that target cancer cells. After purification and expansion in vitro, the CAR T cells are injected into the patient. Although the therapeutic benefits of CAR-T cell therapy should not be overlooked, the limited cell sources, hard-to-be-engineered, time-consuming and money-consuming, heterogeneity of using primary immune cells hinder their applications in cancer therapy.

Combination of PSC differentiation technologies and synthetic CAR may address these challenges. Recently, one of the important examples is CAR iPSC-derived-NK cells [67]. Genetically modified iPSC-derived NK cells hold great potential to offer a safer, standardized and off-the-shelf antitumor therapy. Significantly, early-stage clinical trials of iPSC-derived-NK cells demonstrate safety and efficacy [67].

PSC-derived NK cells were initially differentiated from hESCs, and protocols have been modified to differentiate them from hiPSCs. Such engineered PSC-derived NK cells enable limitless and uniform NK cells derived from certain starting PSC line, and avoid collecting cells from individual donors or cord blood units. Besides, the starting PSC population can be more easily genetically modified than primary NK cells, especially when multiple genetic modifications (CAR and other modifications to enhance functions) are needed [68, 69], and the resulting NK cells derived from those engineered PSCs can express these genetic modifications more uniformly [70]. For example, in vivo persistence of cells is important for long-term therapeutic effects, and activation or inactivation of various genes (IL-15 overexpression or CISH knockout [71]) can enhance the expansion of certain cell types. However, such gene editing may disturb the differentiation process from PSCs to the final cell types (CISH in NK cells is a good example [71]), so in future inducible gene editing is required. PSCs engineered by genetic circuits containing CARs and licensed food additives-induced IL-15 overexpression and/or CISH knockout, may differentiate into NK cells with better persistence in vivo.

In addition, engineered cells with prostate-specific antigen (PSA)-specific GEMS were used to detect pathological PSA levels in the serum of patients with prostate cancer for diagnosis [72].

Dynamically monitoring and eliminating tumorigenic cells

One of the key concerns of hPSC application in cell therapy is the underlying tumorigenicity caused by undifferentiated PSCs and progenitor cells. Therefore, dynamically monitoring and effectively eliminating tumorigenic cells are critical to guarantee the safety of treatment. Engineering cells with inducible suicide or elimination switches can be a useful approach to simultaneously monitor and remove the risk of tumorigenesis.

Scientists reported that miRNA switches could sensitively identify and eliminate undifferentiated cells in iPSC-derived midbrain dopaminergic (mDA)-like neuronal cells [40]. As microRNA-302a-5p (miR-302a) is specifically expressed at high levels in PSCs, the group engineered a miR-302a switch in iPSCs, which functions to activate reporter protein translation when iPSCs differentiate into mDA cells with down-regulation of miR-302a. In SCID mice, miR-302a switch-sorted mDA cells mixed with iPSCs do not undergo teratoma formation. And when the puromycin-resistant gene is under the translational regulation of the miR-302a switch, puromycin-controlled selective elimination of iPSCs can be achieved in vitro.

The iPSC-derived neural stem/progenitor cells (iPSC-derived NS/PCs) are potential for the treatment of SCI, but they lead to tumorigenesis demonstrated by transplantation experiments in NOD/SCID mice. To address this issue, researchers reported the fail-safe systems that were designed based on suicide genes Caspase9 and HSVtk [38, 39]. By dosing NOD/SCID mice with a small molecule CID (AP20817) to induce Caspase9 expression, all engrafted cells in the injured spinal cord could rapidly orient to apoptosis [38]. Since ganciclovir (GCV) can be converted to cell cycle-dependent cytotoxic GCV-triphosphate by HSVtk, unlike the Caspase9/CID system, the HSVtk/GCV system can selectively ablate actively multiplying tumorigenic cells, including iPSCs and iPSC-derived NS/PCs [71]. Transplantation experiments in SCI mice model suggested that the HSVtk/GCV system prevented tumorigenic transformation of engrafted cells and that mature neural cells can preserve for maintaining the improvement of motor function.

In addition, our group also established a safety HLA-A11^R hESC line immune-compatible with more than 20 % of Chinese population by installation of an AP1903-induced suicide switch [73]. AP1903 is the dimerizer agent rimiducid, and it will induce the caspase-9 based suicide system of the modified hESC and lead to cell death [73].

Together, engineering hPSCs with safety-guard circuits and therapeutic circuits might provide huge potential to broaden the application of hPSCs in cell therapy.

Engineering short-range cell-cell communication for self-organization

Synthetic biology can be applied to artificially construct direct or paracrine-mediated intercellular communication (Figure 5). Currently, the synNotch system is the most widely used synthetic receptor to manipulate the intercellular communication mediated by cell contact. This system contains a sender cell, whose cell surface expresses a specific antigen as a ligand; and a receiver cell, which expresses the corresponding synNotch receptor on the cell surface [63, 74]. The synNotch receptor is mainly composed of ECD, TMD and intracellular transcription factor domain. In view of the mechanism on signal transmission of the natural Notch pathway, when the sender cell contacts the receiver cell, the synNotch receptor of the receiver cell will recognize the corresponding antigen, and mediate the cleavage of the intracellular transcription factor domain, initiating the expression of target genes. The synNotch receptor system has various applications, such as achieving specific spatial arrangements and tracing differentiation of organoids. Firstly, by superimposing different synNotch receptor lines, a hierarchical structure of concentric circles [63] or concentric spheres [74] can be achieved in 2D or 3D. Secondly, this system was effectively used for lineage tracing or for reflecting the direct contact of cells during mouse embryonic development as well [75]. Third, engineering downstream responses of the synNotch receptor pathway can stimulate cell-fate decisions. This strategy was applied to mouse ESCs by engineering the synNotch receptor [76]. When Receiver cells are activated, the expression of the neuronal differentiation factor Neurogennin1 is initiated, thereby inducing the differentiation of receiver cells into neurons. Similarly, synNotch can be used to control other differentiation pathways, inducing the production of natural or synthetic morphogen [77], or inducing the expression of adhesion molecule [74] that leads to cell rearrangements.

Figure 5:

Manipulation of self-organization in hPSC-derived organoids via short-range cell-cell communication. Different synthetic cell-cell communication circuits (ligand-receptor pairs) are transferred into several hPSC lines, respectively. These engineered hPSC lines then undergo stepwise differentiation into certain progenitor cell types, respectively, which subsequently self-organize into designer organoids with specific spatial arrangement. The self-organization is directed by the ligand-receptor interaction, activation of receptors, and the activation of target genes including adhesion genes.

Recapitulating morphogen signaling to steer endogenous responses for multicellular system formation

In many developing organs, the spatiotemporal patterns of cellular position and fate decision are directed by graded concentrations of diffusible signal proteins, called morphogens. Cells dynamically regulate corresponding intrinsic GRNs in response to different concentrations of morphogens, obtain respective cell fates, and eventually form complex structure-organized multicellular system. In view of this, engineering stem cells to re-construct morphogen-mediated signaling or GRNs might facilitate the formation of multicellular system (Figure 6). Optogenetic chimeric receptors have been developed to manually control morphogen-mediated signaling or GRNs for mimicking the regulation of morphogen gradients. The basic principle to design optogenetic chimeric receptors is to fuse a photo-inducible domain to cytoplasmic domain of morphogen receptors, stimulating diverse morphogen signaling within mammalian cells, including bone morphogenetic protein (BMP), extracellular signal-regulated kinase (ERK), fibroblast growth factor (FGF), transforming growth factor beta (TGFβ) and Wnt [78], [79], [80]. Optogenetic chimeric receptor, which is constructed by fusing photolyase domain of blue-light photoreceptor Cryptochrome 2 (Cry2) to the cytoplasmic domain of canonical Wnt/β-catenin-response domain lipoprotein receptor-related protein 6 (LRP6), was used to mimic Wnt signaling within subpopulation of hESCs, facilitating hESC differentiation towards mesoderm with tissue-scale self-organized spatial patterns [78], indicating the great potential of generating multicellular system via directly engineering stem cells with synthetic receptors capable of mimicking morphogen signaling (Figure 6).

Figure 6:

Recapitulation of artificial morphogen signaling via synthetic diffusible communication circuits. Type-A sender cell secret artificial morphogen A (green balls), type-B sender cells secret artificial morphogen B (red balls). Receiver hPSCs express both anti-A receptor and anti-B receptor, and the activation of receptors will trigger the expression of their target genes, respectively. Cells in the top sense strong A and weak B, cells in the middle sense moderate A and B, and those in the bottom sense strong B and weak A; and they differentiate into distinct cell types. Finally, multicellular patterning is programmed.

Apart from stimulating morphogen gradient-mediated signaling regulation, direct recapitulation of morphogen gradients could facilitate the construction of a multicellular system via precise control of spatial and temporal regulation. An orthogonal synthetic morphogen system has been developed by reforming synNotch system [77]. This morphogen system can sense soluble GFP secreted by secretor cells, and synergistically form GFP gradient by both anchor cells and anti-GFP receiver cells. Although this system was designed for orthogonal diffusible signal (GFP), GFP can be exchanged with any morphogen as long as a morphogen-specific anchor is available. With the increased number of synthetic receptors recognizing soluble factors, this proof-of-concept study paves the way for broader applications to generate multicellular systems.

Engineering cell-adhesion for multicellular architectures

As mentioned above, in vitro multicellular system formation mainly depends on self-assembling capacity, so high heterogeneity in both shape and cell-types might be due to the random cell self-aggregation instead of tight regulated cell-aggregation as in vivo. Therefore, artificial manipulation of cell-cell adhesion to define tissue-like structure might be another underlying approach to obtain multicellular system with high-fidelity architectures. To date, both natural and orthogonal cell-cell adhesion toolkits have been developed, including synNotch-cadherin, helixCAM and synCAM [74, 81], [82], [83]. Particularly, synNotch-cadherin system has been adapted and verified in mouse ESCs and chimeric mouse embryos [76], suggesting the applications of these cell-cell adhesion toolkits in various human PSC-derived cells for artificial assembling. Besides, these toolkits also exhibit great potential in regulating formation of embryoids, since engineering expression-levels of cell adhesion molecule has been reported to increase the assembling efficiency up to 3-fold of self-organized embryoid, which was formed by aggregation of ESCs, trophoblast stem cells and extraembryonic endoderm [84]. Together, harnessing these toolkits to manipulate structures of multicellular system could promote the self-aggregation efficiency and further obtain the designated organ-structures.

Precise control of multiple-lineage differentiation

Development process is tightly regulated by intracellular GRNs in response to extracellular signals or intracellular transduction. As mentioned above, current strategies to form organoids with multiple lineages are mainly by mixing of different lineages together or optimizing small-molecule cocktail to induce simultaneous differentiation of multiple lineages. However, the broader applications of these approaches are hindered, since cocktail protocol to induce arbitrarily combination of multiple-lineage differentiation, as well as culture medium to support arbitrarily multi-cell mixture are limited and difficult to develop. Recent years, the emerging of abundant omics data, particularly the single-cell transcriptomic profiles from diverse tissues and organs, presents an opportunity to compare gene expression patterns between hPSC-derived organoids and their in vivo counterparts, guiding the design of synthetic genetic circuits for controlling multiple-lineage differentiation and resulting in organoids with enhanced structures and function (Figure 7). One of the proof-of-concept studies demonstrated that human liver organoids with multi-lineages including hepatocyte, biliary, stellate cells and endothelial cells, were generated by using genetic circuits to up-regulate PROX1, ATF5 and CRISPR-based activation of endogenous CYP3A4, respectively [48].

Figure 7:

Development of designer hPSC-derived organoids directed by synthetic genetic circuits. Comparison of expression profile between organoids and their native counterparts reveals several candidate TFs. Based on the TF candidates and their expression pattern, synthetic genetic circuits are designed, built, and transferred into hPSCs, which will direct the generation designer organoids with improved structures and functions. TFs, transcription factors.