Induced pluripotent stem cells (iPSCs): molecular mechanisms of induction and applications

Abstract

The induced pluripotent stem cell (iPSC) technology has transformed in vitro research and holds great promise to advance regenerative medicine. iPSCs have the capacity for an almost unlimited expansion, are amenable to genetic engineering, and can be differentiated into most somatic cell types. iPSCs have been widely applied to model human development and diseases, perform drug screening, and develop cell therapies. In this review, we outline key developments in the iPSC field and highlight the immense versatility of the iPSC technology for in vitro modeling and therapeutic applications. We begin by discussing the pivotal discoveries that revealed the potential of a somatic cell nucleus for reprogramming and led to successful generation of iPSCs. We consider the molecular mechanisms and dynamics of somatic cell reprogramming as well as the numerous methods available to induce pluripotency. Subsequently, we discuss various iPSC-based cellular models, from mono-cultures of a single cell type to complex three-dimensional organoids, and how these models can be applied to elucidate the mechanisms of human development and diseases. We use examples of neurological disorders, coronavirus disease 2019 (COVID-19), and cancer to highlight the diversity of disease-specific phenotypes that can be modeled using iPSC-derived cells. We also consider how iPSC-derived cellular models can be used in high-throughput drug screening and drug toxicity studies. Finally, we discuss the process of developing autologous and allogeneic iPSC-based cell therapies and their potential to alleviate human diseases.

Fig. 1

Development of the induced pluripotent stem cell (iPSC) technology. a A timeline of key breakthroughs related to the iPSC technology. b (Top) Somatic cell nuclear transfer (SCNT) experiments were pioneered by John Gurdon in the African clawed frog. Gurdon demonstrated that somatic cells retained all the genetic information necessary to give rise to a germline-competent organism. Successful SCNT in mammals was demonstrated by Keith Campbell, Ian Wilmut, and colleagues who cloned Dolly the sheep. (Bottom) Masako Tada and colleagues demonstrated that pluripotency can also be achieved by fusing a somatic cell with an embryonic stem cell, leading to the formation of a hybrid tetraploid cell. 4N, tetraploid. c The groundbreaking experiments of fibroblast reprogramming to pluripotency were pioneered by Kazutoshi Takahashi and Shinya Yamanaka. The researchers selected 24 factors as candidates for reprogramming and delivered these factors into mouse fibroblasts in various combinations by retroviral transduction. Eventually, Takahashi and Yamanaka identified a combination of 4 reprogramming factors—Oct4, Sox2, Klf4, and Myc—that was sufficient to reprogram mouse fibroblasts into embryonic stem cell-like pluripotent cells, known as iPSCs. Subsequently, Yamanaka and James Thomson independently reprogrammed human fibroblasts into iPSCs in 2007

Fig. 2

Induced pluripotent stem cell (iPSC)-derived cellular models. The iPSC technology can be applied to derive cellular models of varying complexity, ranging from two-dimensional mono-cultures to three-dimensional multicellular assemblies. Various neural cellular models are shown as an example. a Differentiation of neural progenitor cells (NPCs) from iPSCs is achieved by promoting neuroectoderm specification by dual SMAD inhibition. Subsequently, NPCs can be differentiated into terminal neural lineage cells, such as neurons and astrocytes. b iPSC-derived cells can be maintained in a mono-culture or together with other cell types in a co-culture. Different cell types can also be assembled into an organ-on-a-chip that contains separate compartments and enables modeling of complex tissue architecture. Alternatively, iPSC-derived cells can be transplanted in vivo to expose the cells to a complex tissue environment. c iPSCs can be differentiated into three-dimensional self-organizing organoids that partially resemble endogenous tissue architecture and contain several cells types. Organoids can also be transplanted in vivo to promote their vascularization and maturation. d Different types of organoids can be fused together into assembloids for the study of higher-order tissue interactions, such as long-distance innervation and cell migration

Fig. 3

Disease modeling with iPSC-derived cells. a Genetic diseases, such as Alexander disease (AxD), can be modeled using patient-derived iPSCs that carry disease-causing mutations. A tissue biopsy is first taken from a patient with AxD. Somatic cells are reprogrammed into iPSCs, and the GFAP mutations that cause AxD are corrected by gene editing. Patient-derived iPSCs and isogenic corrected controls are then differentiated into astrocytes that express GFAP at high levels. Co-culture of AxD astrocytes with oligodendrocyte progenitor cells (OPCs) reveals impaired OPC proliferation and oligodendrocyte (OL) myelination. Transcriptomic analysis indicates increased expression of the CHI3L1 gene, whereas OPC dysfunction can be partially reversed by CHI3L1 protein depletion. These observations in vitro can be further validated in primary human brain tissues as well as and experiments in vivo. b Sporadic diseases, such as Alzheimer’s disease (AD), can be modeled with patient-derived iPSCs that harbor genetic risk factors; alternatively, iPSC-derived cells can be exposed to non-genetic risk factors to induce disease-relevant pathology. For example, exposure of iPSC-derived brain organoids to human serum mimics the breakdown of the blood-brain barrier and induces AD-like pathology. Brain organoids exposed to neurotoxic serum factors have increased levels of toxic amyloid peptides and hyperphosphorylated tau as well as exhibit impaired neuronal activity. c Infectious diseases, such as COVID-19, can be modeled by exposing iPSC-derived cells and organoids to viral pathogens. iPSC-based models of viral infection can reveal human-specific tropism, mechanisms of entry, and other features of a particular virus

Fig. 4

Modeling aging-associated phenotypes with iPSC-derived cells. One important limitation of using iPSC-derived cells to model human diseases is their fetal-like phenotypes and the lack of aging-associated cellular features. The process of somatic cell reprogramming to iPSCs is associated with a nearly complete erasure of aging-associated epigenetic marks and phenotypes. Therefore, various strategies to induce aging-associated phenotypes in iPSC-derived cells have been developed. a Exposure of iPSC-derived cells to compounds that disrupt cellular homeostasis can be used to induce aging-associated phenotypes, such as mitochondrial stress or cellular senescence. For example, rotenone disrupts electron transfer in mitochondria, leading to an increased production of reactive oxygen species that can cause mitochondrial stress, damage other organelles, and induce cellular senescence. b Aging-associated phenotypes can also be induced by ectopic expression of progerin, a truncated variant of lamin A nuclear lamina protein. Progerin causes the Hutchinson-Gilford progeria syndrome, a disease that manifests as accelerated aging due to the disruption of the nuclear lamina. Ectopic expression of progerin is sufficient to induce senescence- and aging-associated phenotypes in iPSC-derived neurons and other cells. c Aging-associated phenotypes are preserved if target cells are derived by direct transdifferentiation without an iPSC intermediate. Primary fibroblasts can be transdifferentiated into neurons that exhibit aging-associated phenotypes and epigenetic age signatures of the fibroblast donor, and can thus be used to study age-related dysfunction of neural cells

Fig. 5

Autologous and allogeneic iPSC-based cell therapy. In autologous cell therapy, somatic cells are collected from the patient who will receive the cell transplant. The isolated somatic cells are reprogrammed into iPSCs, which can then be genetically engineered to correct disease-associated mutations or introduce new gene expression vectors. Modified iPSCs are differentiated into the cellular product that will be transplanted into the patient and rigorously evaluated for quality. In allogeneic cell therapy, iPSCs are taken from a biobank and genetically engineered for immune cloaking. The resulting hypoimmunogenic universal donor iPSCs can be further genetically modified to introduce cell therapy-specific gene expression vectors, such as a chimeric antigen receptor (CAR) expression cassette, and then differentiated into the desired cell type. After rigorous quality assessment, cellular products can be stocked and distributed as off-the-shelf therapeutics for transplantation into multiple recipients. KO, knockout; KI, knockin

Fig. 6

Development of iPSC-based autologous cell therapy. Despite the success of adoptive immune cell therapy, multiple other diseases affect cell types that cannot be easily isolated from patients for genetic engineering and transplantation back into the patient. For example, Canavan disease (CD) is a monogenic autosomal recessive neurological disorder caused by mutations in the aspartoacylase (ASPA) gene. These mutations disrupt ASPA enzymatic activity, leading to the accumulation of N-acetylaspartate (NAA) in the brain and causing spongy degeneration. ASPA enzymatic activity can be restored by transplantation of autologous neural progenitor cells (NPCs) that harbor CRISPR/Cas9-corrected ASPA or ectopically express wild-type ASPA delivered by lentiviral (LV) transduction. a A skin biopsy is obtained from a CD patient, and patient-specific iPSCs are derived from the isolated skin fibroblasts. b iPSCs are genetically engineered to restore wild-type ASPA expression and differentiated into NPCs that will be used for transplantation. c To demonstrate the efficacy of iPSC-derived NPC therapy for CD, preclinical experiments using a CD mouse model (Nur7) can be performed. CD mice exhibit characteristic spongy degeneration with vacuolation, myelin defects, and motor dysfunction. In our studies,^,, we transplanted WT-ASPA-NPCs into the corpus callosum (CC), the subcortical region (SC), and the brainstem (BS) by stereotactic injection. We found that WT-ASPA-NPC-transplanted CD mice exhibited increased ASPA activity and reduced NAA levels, increased myelination and reduced vacuolation, and improved motor function. GMP, good manufacturing practice

Fig. 7

Engineering universal donor cells for allogeneic cell therapy. a Universal donor cells are genetically engineered to prevent the host immune response despite their foreign origin. CD8⁺ cytotoxic T cells recognize foreign cells via their T cell receptor (TCR) that interacts with human leukocyte antigen (HLA) class I molecules presenting unique antigens. If a foreign antigen is presented by the HLA class I molecules, CD8⁺ cytotoxic T cells initiate destruction of the encountered cell. Knockout of the β₂ microglobulin (B2M) gene is sufficient to disrupt the universal donor cell interaction with CD8⁺ cytotoxic T cells. However, ablation of HLA class I molecules elicits a “missing-self” response by natural killer (NK) cells, leading to cell lysis. Therefore, B2M knockout is often combined with ectopic expression of HLA-E, which interacts with the inhibitory NK cell receptor NKG2A/CD94 to suppress the missing-self response. Knockout of CIITA disrupts foreign antigen presentation to CD4⁺ T cells via HLA class II molecules. To prevent macrophage-mediated cell killing, CD47 surface protein can be ectopically expressed in universal donor cells. CD47 interacts with the signal-regulatory protein α (SIRPα) and acts as the “don’t-eat-me” signal to suppress macrophage-mediated phagocytosis. b Hypoimmunogenicity of universal donor cells can be tested by performing in vitro and in vivo cytotoxicity assays, in which universal donor cells are mixed with primary immune cells, such as T cells, derived from an unrelated donor. Universal donor cells exhibit increased survival and stable persistence in the presence of primary immune cells of a mismatched donor, indicating successful immune evasion