Tuning cell fate

Abstract

Epigenetic interventions are required to induce reprogramming from one cell type to another. At present, various cellular reprogramming methods such as somatic cell nuclear transfer, cell fusion, and direct reprogramming using transcription factors have been reported. In particular, direct reprogramming from somatic cells to induced pluripotent stem cells (iPSCs) has been achieved using defined factors that play important epigenetic roles. Although the mechanisms underlying cellular reprogramming and vertebrate regeneration, including appendage regeneration, remain unknown, dedifferentiation occurs at an early phase in both the events, and both events are contrasting with regard to cell death. We compared the current status of changes in cell fate of iPSCs with that of vertebrate regeneration and suggested that substantial insights into vertebrate regeneration should be helpful for safe applications of iPSCs to medicine.

Introduction

Newt limbs, zebrafish fins, and xenopus tails can regenerate after amputation. The phenomena called appendage regeneration have been intensively studied; however, a gap between this regeneration and mammalian regenerative biology still remains.^1-3 With regard to limb regeneration in salamanders, which is a representative example of appendage regeneration, amputated limbs are covered by the epidermis, and immature cells accumulate and proliferate beneath them, forming the “blastema.”⁴ The blastema, which is encased by the newly formed epidermis, comprises dedifferentiated cells that are restricted to skeletal muscle cells, chondrocytes, Schwann cells, and mesenchymal fibroblasts.⁵ Salamanders and zebrafish can also regenerate cardiac tissue after the resection of the ventricular apex^6,7 or the genetic destruction of cardiomyocytes.⁸ Significant cardiomyogenesis was observed in murine hearts less than 1 wk after birth.⁹ The vertebrate regeneration discussed above entails a common process, dedifferentiation. However, the factors distinguishing regenerative vertebrates from nonregenerative vertebrates remain unknown.¹⁰ Thus, in regenerative biology and medicine, it is crucial to understand the molecular mechanisms underlying the processes, including growth, patterning, dedifferentiation, and redifferentiation.

Appendage Regeneration

Overview

In zebrafish fin regeneration after amputation, there are four stages, termed “regeneration epithelialization or wound healing” (0–1 d post-amputation [dpa]), “blastema formation” (1–2 dpa), “regenerative outgrowth” (2–7 dpa), and “termination.”^11,12 At 1 dpa, the proximal epidermis migrates to cover the stump and form a 3–4-cell-thick layer. Inflammation then proceeds to clean the clotted plasma and cellular debris. At 2 dpa, histolysis occurs, remodeling the extracellular matrix, and cells to be liberated by histolysis begin to dedifferentiate in this period. They do not significantly activate the cell cycle and primarily form the blastema by distal migration under the wound epidermis.¹³ Subsequently, the accumulation blastema is achieved by a marked increase in mitosis, which is dependent on factors from the wound epidermis¹⁴ and regenerating nerve.⁴ Subsequent regenerative outgrowth or blastema accumulation is characterized by robust proliferation of dedifferentiated cells. An outstanding feature is the presence of rare apoptotic cells in the blastema despite the presence of avascular tissue, which could be hypoxic and possibly susceptible to apoptosis.

Proteomic analysis of the blastema in regenerating axolotl limbs revealed that dedifferentiated cells are capable of avoiding apoptosis through several mechanisms such as reduced metabolism, differential regulation of proapoptotic and antiapoptotic proteins, and initiation of an unfolded protein response.¹² The tricarboxylic acid cycle and electron transport enzymes are downregulated by 7 dpa, which may be mediated via increased nitrogen oxide in the blastema. The proteasome deliberately destroys aberrant proteins that could not have been refolded by a chaperon and those that have been specifically marked by a recognition tag formed from ubiquitin. In the axolotl limb regeneration blastema, several chaperons and components of the proteasome–ubiquitin system are upregulated at an early phase, suggesting that protein quality control could play an essential role in dedifferentiation.¹²

Molecular mechanism

Loss-of-function experiments have revealed regulators of the regeneration process,^15-17 although further elucidation is required. Till date, canonical Wnt and fibroblast growth factor (FGF) signaling pathways are the major pathways to be reported in appendage regeneration, and activin, retinoic acid, hedgehog, and noncanonical Wnt signaling is involved in the regulation of regeneration.¹¹ Moreover, as negative regulators, noncoding microRNAs play an important role by downregulating the expression of ligands for FGF and Wnt/β-catenin signaling. Lef1, a ligand for canonical Wnt, is expressed in the newly developed epithelium over the amputated plane. Inducible DKK1, an inhibitor of canonical Wnt signaling, inhibits lef1 expression and blastema formation, and the knockdown following this stage decreases the sizes of regenerated fins.¹⁵ Although the lateral epithelial growth that wraps the amputation plane could not be prevented by DKK1 during 0–1 dpa, msxb expression, which is involved in FGF signaling, was impaired. The specific roles of canonical Wnt/β-catenin molecules for each regeneration stage and crosstalk with the FGF signaling pathway have been uncovered.

In contrast, Lef1 expression in the developed epithelium leads to fgfr1 expression in mesenchymal cells of the blastema. Chemical impairment of fgfr1 prevented blastema formation, msxb expression, and consequently cell proliferation.¹⁶ Moreover, blastema formation was blocked in fgf20a null mutants.¹⁷ One example of the negative impact to Wnt/β-catenin signaling is Wnt5b, a noncanonical Wnt.¹⁵ In addition, miR-203 represses lef1 expression as a mediator of Wnt/β-catenin.¹⁸ In contrast, miR-133 is involved in the FGF signaling pathway.¹⁹ An array of noncoding RNAs should form a sophisticated regulatory network for appendage regeneration, which may share features with the regulatory network for carcinogenesis.

Comprehensive transcriptional profiling²⁰ and RNA sequencing²¹ during limb regeneration revealed significant upregulation of c-Myc (myelocytomatosis oncogene) and Krüppel-like factors 4 (Klf4). In contrast Oct4 (POU domain, class 5, transcriptional factor 1), Sox2 (sex-determining region Y-box 2), and Nanog were not upregulated. SALL4, which is involved in the maintenance of pluripotency, was overexpressed during blastema formation. In contrast, SALL1 and SALL3 were only gradually expressed during the patterning phase. During epithelialization and blastema formation, the expression of many oncogenes such as ATF3, JUN3, EGR1, NR4A2, and FOS increased; however, these genes were then downregulated during the patterning process. Proteomic analysis of the blastema in regenerating axolotl limbs showed upregulation of LIN28, which is related to cellular reprogramming. Also, antiapoptotic mechanisms, such as reduced metabolism and initiation of an unfolded protein response were activated.¹²

Cell sources

Determination of the origin of blastema cells has been one of the main concerns of regenerative biology for a long time.¹ Genetic lineage tracing elucidated the origin and differentiation capability of blastema cells in amputated axolotl limbs and zebrafish fins. Transplantation experiment using green fluorescent protein (GFP) cells from various tissues of axolotol limbs showed that grafted cells dedifferentiate, proliferate, and redifferentiate into cells that are restricted to the origin.⁵ Cre/loxP-based genetic marking to track osteoblasts in zebrafish fin regeneration clearly demonstrated the dedifferentiation of pre-existing osteoblasts and redifferentiation to osteoblasts.²² However, it is possible that resident stem cells are involved in appendage regeneration,²³ particularly in the case of skeletal muscles, which are accompanied by a population of stem cells called satellite cells.²⁴ Moreover, genetic ablation of all skeletal osteoblasts in zebrafish fins resulted in de novo osteogenetic process, rather than through the dedifferentiation and redifferentiation process.²⁵

Vertebrate Regeneration in the Heart

The zebrafish heart has been intensively investigated for its regenerative capacity, and amputation experiments have recently provided convincing proof of the dedifferentiation model in vertebrate regeneration using genetic fate mapping.^7,26 Genetic cell ablation by inducible diphtheria toxin expression in zebrafish hearts has strengthened the evidence that newly developed cardiomyocytes are derived from pre-existing cardiomyocytes through a dedifferentiation process.⁸ Msp1, which is a mitotic checkpoint kinase, and GATA4, which is a transcription factor (TF) of early cardiac development, are essential for zebrafish cardiac regeneration.^7,27 Similar to limb regeneration, three phases of cardiac apex regeneration have been distinguished as “inflammatory,” “reparative,” and “regenerative,” and these correspond to wound healing, blastema formation, and outgrowth and termination phases in limb regeneration, respectively.²⁸ With dedifferentiation following apex amputation, epicardiac cells proceed into the epithelial–mesenchymal transition (EMT) in response to FGF and PDGF.^29,30 Thereafter, cardiomyocytes with disorganized sarcomeres are similar to immature cardiomyocytes that have been derived from induced pluripotent stem cells (iPSCs) using current cardiac differentiation protocols as monolayers or using embryonic body methods. These detach from one another and launch proliferation with the expression of positive cell cycle regulators such as polo-like kinase 1 (plk1) and cdc2.²⁶ Within 1 wk of birth, neonatal mice repopulate amputated cardiac apexes with newly developed cardiomyocytes,⁹ which are formed through dedifferentiation and redifferentiation, a phenomenon similar to that observed in zebrafish hearts. This ability was found to be lost by 1 wk of age, and injured apexes were then filled with fibrotic tissues (Fig. 1).

Figure 1. Vertebrate regeneration in mouse neonates hearts.

Reprogramming into Pluripotency

During development, gene expression is regulated by TF and epigenomic networks. One method of determining whether the gene regulatory mechanisms can be reversed is somatic cell nuclear transfer (SCNT). The concept of nuclear transfer³¹ originates from proposals by Hans Spemann in 1936, who was motivated to test Weismann’s theory that cell differentiation is the consequence of unequal nuclear division during embryonic development. In 1952, Briggs and King successfully accomplished nuclear transplantation of a cell from the blastula into frog eggs for the first time.³² In 1958, John Gurdon demonstrated that nuclei from intestinal epithelial cells could be developed to produce offspring when transferred into enucleated eggs.³³ In 1997, Ian Willmut verified that SCNT was successful in sheep.³⁴ Shinya Yamanaka revealed a set of genes to drive reprogramming to pluripotency in 2006.³⁵ Until the discovery of iPSCs, the differentiation process was considered to be one-way, with the exception of the fertilization process. This process of pluripotent stem cell (PSC) differentiation has been shown to be regulated and maintained by complex transcriptional and epigenetic networks.^36-38 According to the classical view of cell fate hierarchy based on the Waddington epigenetic landscape,³⁹ PSCs reside at the top of the hierarchy above differentiated somatic cells (Fig. 2). This model indicates a natural restriction of the cell differentiation potential during normal cellular development along each lineage.

Figure 2. Scheme of cell fate changes based on Waddington’s epigenetic landscape. Direct reprogramming is the reversion of terminally differentiated cells such as fibroblasts to a pluripotent state. Direct conversion is the alteration from one cell type to another, such as fibroblasts to cardiomyocytes. Dedifferentiation is defined as a reversion of specialized phenotypes into an undifferentiated state. Transdetermination is the switching of somatic stem/progenitor cells from one determined state to another closely related state.^110,111

Three cellular reprogramming methods for converting somatic cells to PSCs have been reported, including SCNT, cell fusion, and direct reprogramming using TFs (Fig. 3). Briggs and King²⁹ reported the first cellular reprogramming by transplanting intact nuclei from amphibian blastocysts into enucleated oocytes using SCNT. The transferred oocyte nuclei were activated to zygotes, which developed into tadpoles. In addition, John Gurdon, Nobel laureate in 2012, challenged SCNT using more differentiated nuclei from a tadpole intestinal cell and succeeded in producing offspring.³³ These results indicate that genomic DNA from differentiated cells remains unchanged and retains the capacity to develop reproductive organisms. Although the efficiency of successful cloning is only 1–2%,⁴⁰ some reports demonstrate cloning of mammals such as sheep⁴¹ and mice⁴² from nuclei of adult cells. Recently, Tachibana et al.⁴³ reported the reprogramming of human somatic cells to pluripotent cells using SCNT. However, the requirement of skillful handling, ethical issues, limited oocyte availability, and low success rates hamper this approach for studying reprogramming mechanisms.

Figure 3. Study design of nuclear reprogramming Transcriptional factor (TF)-mediated reprogramming (upper left); reprogramming TFs such as Oct4, Sox2, Klf4, and c-Myc (OSKM) are introduced into iPSCs from somatic cells. Cell fusion (upper right); two or more different types of cells are fused using methods such as electrical cell fusion, polyethylene glycol cell fusion, or Sendai virus-induced cell fusion. Somatic cell nuclear transfer (SCNT; lower); nuclei from donor cells are transferred into enucleated oocytes. Blastocysts derived from SCNT-oocytes can then be cultured as nuclear transfer-embryonic stem cells (NT-ESCs) or can be implanted into pseudopregnant mice to produce offspring.

Another cellular reprogramming method uses human ESCs to reprogram myeloid precursors following cell–cell fusion.⁴⁴ Several methods promote cell fusion, including the use of polyethylene glycol, Sendai viruses, and electric pulses. Cell fusion using human or mouse ESCs offers an inexpensive and accessible system, in which the sequence of remodeling events results in the successful conversion of somatic cells to an ES-like state.^45-49 This approach provided essential insights into the mechanisms underlying the reversion of cell fate, including epigenetic plasticity. However, cell fusions of two or more different somatic cells led to the formation of heterokaryons, in which parental nuclei remain discrete within the same cytoplasm and are therefore unsuitable for medical applications such as in regenerative medicine.

The hypothesis that reprogramming factors exist in the cytoplasm, as suggested by SCNT,^32,33,43 ESC fusion experiments,^47,49,50 and myogenic differentiation using overexpressed MyoD,⁵¹ open a door into the frontier of TF-mediated reprogramming. Yamanaka et al.^43,44 generated PSCs from somatic cells using TFs.^35,52 The four TFs, Oct4, Sox2, Klf4, and c-Myc (referred to as OSKM in this review), were sufficient to reprogram somatic cells into pluripotent cells known as iPSCs.^53-56 At present, many studies have reported alternative factors such as TFs,⁵⁵ small molecules,^57,58 microRNAs,^59,60 proteins,⁶¹ mRNAs,⁶² and plasmids^63,64 that can accelerate and increase the reprogramming state.

Mechanisms in Reprogramming during the Initiation Phase

To obtain mechanistic insights into reprogramming, various approaches, including proteomics, comprehensive microarray analysis, metabolomics, and single cell technologies, have been applied to the process of TF-mediated reprogramming to iPSCs. The reprogramming process is delineated into three phases on the basis of the gene expression status. These are known as “initiation,” “maturation,” and “stabilization.”⁶⁵ Changes in both mRNA and microRNA expression as well as histone modification result in two big waves during the initiation phase and at the start of the stabilization phase, known as the first and second waves, respectively.⁵⁴

During the initiation phase, donor cells promptly initiate MET,^55,66 robust cell proliferation, metabolic changes, and alternative histone modifications. Immediately after exogenous OSKM expression, MK (c-Myc, Klf4) proteins attach to the promoters of genes with open chromatin and proceed to either activate or repress depending on the type of the downstream gene, and OSK (Oct4, Sox2, Klf4) proteins bind multiple distal enhancer regions of genes, except those involved in maintaining pluripotency, which they do not occupy as seen in the case of PSCs.⁶⁷ Therefore, OSKs are known as promiscuous pioneer factors.⁶⁸ Such promiscuous binding facilitates the de novo accumulation of histone H3 that is dimethylated at lysine 4 (H3K4me2) on early genes of the reprogramming process such as F-box only protein 15 (Fbxo15), Fgf4, and Sall4; somatic genes such as those involved in the cell cycle (Cdc20 and Cdc25c); and metabolic genes such as phosphofructokinase liver B-type (pfkl) and glucose phosphate isomerase (Gpi). In contrast, binding of MK proteins to the promoter regions of fibroblast-specific genes, such as thymus cell antigen 1 (Thy-1) and collagen type V α 2 (Col5a2), represses transcription and causes loss of pre-existing H3K4me2. The first wave of changes in gene expression (both increase and decrease) is primarily regulated by c-MYC, and the leading part of the second wave is mediated by OCT4 and SOX2. Single cell quantitative analysis unveiled this heterogeneity of expression in a subset of pluripotency genes.⁵⁵ However, because of the stochastic nature of reprogramming, no predictable specific marker was found for cells that were poised to become bona fide (truly reprogrammed iPSCs) iPSCs prior to the maturation phase.⁶⁹

Evolving Artificial Reprograming of Cell Fate

Dedifferentiation and Redifferentiation

A newly developed strategy for cell fate switching is similar to a physiological process of appendage regeneration, which was attained by sequential treatments using ectopic OSKM expression for short periods; inhibition of JAK/STAT signaling, involving the maintenance of pluripotency; and exposure to the cardiac differentiation culture condition.⁷⁰ OSKM was putatively assumed to induce the dedifferentiation of donor cells into the intermediate state. Thereafter, the cells were intended to be driven into cardiac lineages rather than being launched into the deterministic phase for reprogramming to pluripotency. Because this “primed conversion”⁷¹ operates on a scheme similar to appendage regeneration, with initial dedifferentiation and subsequent redifferentiation, it may be attractive to develop a medical treatment in humans.

We also examined this primed conversion to cardiomyocytes. Subsequently, mouse embryonic fibroblasts (MEFs) were transduced with a cardiomyocyte-specific myosin heavy chain (Myh6) promoter to drive GFP expression. Fourteen days after infection, OSKM-induced MEFs (OSKM-MEFs) locally formed few cell clusters, which were GFP positive and automatically contractile (Fig. 4A). The genes Gata4, Tbx5, Nkx2.5, Baf60c, and Isl1 were strongly expressed; Myh6, Myh7, Mlc2a, and Mlc2v were also expressed, with the strongest being Mlc2a expression. However, Mef2c, Anp, and Bnp expression was undetectable by reverse transcription polymerase chain reaction (RT-PCR) (Fig. 4B). These results indicated that OSKM-MEFs could be heterogeneous and may have included atrial- or ventricle-like cardiomyocytes. No Oct4-positive cells were detected in the cluster (Fig. 4C). However, OSKM-MEF cell clusters were positive for the cardiac-related proteins Gata4 and α-actinin in immunohistochemical experiments (Fig. 4C). To the best of our knowledge, no primed conversion to human cardiomyocytes has been reported.

Figure 4. Cardiac differentiation of OSKM-mediated mouse embryonic fibroblasts (MEFs) via primed conversion (A) Phase contrast microscopic view of OSKM-mediated MEFs 14 d after OSKM infection. OSKM-mediated MEFs were differentiated into cardiomyocytes. (B) Reverse transcription polymerase chain reaction (RT-PCR) analysis of gene expression. (C) Fluorescent microscopic view with immunofluorescent staining of OSKM-mediated MEFs, which were stained with DAPI, OCT4, Gata4, and α-actinin antibodies.

Direct Conversion

Cellular dedifferentiation switches a program that drives the specific function of a somatic cell to another program for either proliferation, cell death, or redifferentiation, irrespective of whether the destination is the same as the origin.⁷² The last avenue without mitosis refers to the conversion. Since the discovery of human iPSCs, regenerative medicine using iPSC-derived differentiated cells has been an enthusiastic target in all areas, particularly cardiology and neurology. The drawbacks of this strategy include the risk of teratoma and time-consuming establishment of bona fide iPSCs that lead to target cells. The direct reprogramming approach provides an additional possibility of differentiation to functional cells such as cardiomyocytes,^73-78 hepatocytes,^79,80 β-cells,⁸⁰ brown fat,⁸¹ chondrocytes,^82,83 chondrosarcoma,⁸⁴ macrophages,^85,86 blood progenitors,⁸⁷ myocytes,⁵¹ and neurons^88-97 using lineage-specific combinations of TFs (Table 1). These methods were known as “direct conversion” (Fig. 3).⁷¹ Cardiomyocyte-like cells (iCMs) were obtained by introducing Gata4, Mef2c, and Tbx5 (GMT) into murine fibroblasts, despite the rare efficacy.⁷⁴ Thereafter, a combination of miRNAs 1, 133, 208, and 499 also induced murine iCMs from fibroblasts in vitro.⁷⁶ Remarkably, in vivo direct gene transfer of GMT⁹⁸ or GMT and HAND2⁷⁵ resulted in the conversion of cardiac fibroblasts to the cardiomyocyte phenotype and attenuated the cardiac function of the infarcted heart in mice. Recently, human fibroblasts were reprogrammed to iCMs with GMT plus ESRRG and MESP1, and ZFPM2 and myocardin enhanced the conversion efficacy.⁹⁹ This strategy may offer some advantages such as a reduced cancer risk and shorter time. However, depending on the desired target cells, the factors used to convert them have to be defined for each. Direct conversion has no corresponding physiological process in nature, and it is a purely artificial interventional strategy. Thus, investigations of mechanisms and safety issues, including long-term follow-up prior to any clinical application, are critical.

Table 1. Summary of previously reported cell state conversions.

Germ layer	Target cell	Species	Parent cell	Inducer	Reference
Mesoderm	Brown fat	H, M	Myoblast	PRDM16, C/EBPβ	81
	Cardiomyocyte	M	Amniotic mesodermal cell	Tbx5, Gata4, Baf60c	73
	Cardiomyocyte	M	Cardiac fibroblast	Tbx5, Gata4, Mef2c	74
	Cardiomyocyte	M	Fibroblast	Tbx5, Gata4, Mef2c, Hand2	75
	Cardiomyocyte	M	Fibroblast	miR-1, miR-133, miR-208, miR-499	76
	Cardiomyocyte	H	Fibroblast	TBX5, GATA4, HAND2, MYOCARDIN, miR-1, miR-133	77
	Cardiomyocyte	H	Fibroblast	TBX5, GATA4, MEF2C, MYOCARDIN, ESRRG, MESP1, ZFPM2	78
	Chondrosarcoma	H	Placental cell	T, BCL6, c-MYC, MITF, BAF60C	84
	Chondrocyte	H, M	Adult dermal fibroblast	SOX9, c-MYC, KLF4	82,83
	Macrophage	M	Lymphoid precursor	C/EBPα, PU.1	85
	Macrophage	M	β-cell	C/EBPα	86
	Multilineage blood progenitors	H	Fibroblast	OCT4, Cytokines	87
	Myocyte	M	Fibroblast	MyoD	51
Endoderm	Hepatocyte	M	Fibroblast	Hnf4α, Foxa1, Foxa2, or Foxa3	79
	Hepatocyte-like cell	M	Fibroblast	Gata4, Hnf1α, Foxa3, and inactivation of p19 (Arf)	80
	β-cell	M	Adult pancreatic exocrine cell	Ngn3, Pdx1, Mafa	80
Ectoderm	Dopaminergic neurons	H, M	Fibroblast	ASCL1, NR4A2, LMX1A	88
	Dopaminergic neurons	H	Fibroblast	ASCL1, BRN2, MYT1L, LMX1A, FOXA2	89
	Functional neuron-like cells	H	Fibroblast	ASCL1, NGN2, small molecule	90
	Functional spinal motor neuron	H, M	Fibroblast	Hb9, ISL1, LHX3, ASCL1, BRN2, MYT1L	91
	Neuron	M	Embryonic fibroblast	Ascl1, Brn2, Myt1l	92
	Neuron	H	Fibroblast	ASCL1, BRN2, MYT1L, NEUROD1	93
	Neuron	H	Fibroblast	MYT1L, BRN2, miR-124 in differentiation medium	94
	Neuron	H	Fibroblast	ASCL1, MYT1L, miR-9/9*, miR-124	95
	Neuronal cell	H	Fibroblast	ASCL1, BRN2, MYT1L, OLIG2, ZIC1	96
	Tripotent neural progenitor	M	Fibroblast	Brn2, Sox2, FoxG1	97

H, Human; M, Mouse.

Perspective

While dedifferentiation is commonly associated with reentry into the cell cycle, its true nature is to withdraw from a given differentiated state.¹⁰⁰ During direct reprogramming to pluripotency, dedifferentiation of donor cells may occur prior to reentry into the cell cycle.¹⁰¹ The fact that supports this notion is that mature B cells need to be dedifferentiated with either C/EBPα or PAX5 before they can be reprogrammed.¹⁰² On the other hand, there are few apoptotic cells during blastema formation,¹⁰³ while many cells that receive OSKM for iPSCs undergo apoptosis.

Because there is no appropriate experimental system to uncouple the stage of dedifferentiation from the cell cycle in animals. One clear example of dedifferentiation is plant protoplasts, which are acquired by the treatment of leaves with cellulase and are a stem cell-like state.⁷² In tobacco (Nicotiana tabacum), the transition to protoplasts (dedifferentiation) is accompanied with the activation of the transposable element Tnt1.¹⁰⁴ In Arabidopsis, the dedifferentiation is associated with large-scale decondensation of the pericentric heterochromatin,¹⁰⁵ related to telomerase-independent telomere lengthening, which often involves DNA recombination.¹⁰⁶ The increased frequency of somatic recombination in Arabidopsis due to infection and environmental stress^107,108 may suggest genetic variation and genome instability during cellular dedifferentiation. The human genome includes transposons and transposon-like repetitive elements, only a small proportion of which remains active. These facts bring attention to the potential risk associated with dedifferentiation processes.¹⁰⁹

Unlike dedifferentiation due to artificial intervention such as TF gene transfer and nuclear transplantation, physiological dedifferentiation in amputational regeneration seems to be well organized and regulated to reform tissues. The difference may be an essential cue to develop regenerative medicine based on reprogramming technology. The validation of harmful genetic variation in human iPSCs has been started through whole genome sequencing. Even if a negative impact is detected using the current technology, a thorough evaluation of appendage regeneration should result in progression in the field of medicine using iPSCs.

Glossary

Abbreviations:

iPSCs

induced pluripotent stem cells

dpa

day post-amputation

FGF

fibroblast growth factor

c-Myc

myelocytomatosis oncogene

Klf4

Krüppel-like factors 4

Oct4 (POU domain

class 5, transcription factor 1

Sox2

sex-determining region Y-box 2

GFP

green fluorescent protein

TF

transcription factor

EMT

epithelial–mesenchymal transition

plk1

polo-like kinase 1

SCNT

somatic cell nuclear transfer

PSC

pluripotent stem cell

OSKM

Oct4, Sox2, Klf4, and c-Myc

MK

c-Myc, Klf4

OSK

Oct4, Sox2, Klf4

H3K4me2

histone H3 that is dimethylated at lysine 4

Fbxo15

F-box only protein 15

pfkl

phosphofructokinase liver B-type

Gpi

glucose phosphate isomerase

Thy-1

thymus cell antigen 1

Col5a2

collagen type V alpha 2

MEFs

mouse embryonic fibroblasts

Myh6

myosin heavy chain

RT-PCR

reverse transcription polymerase chain reaction

iCMs

Cardiomyocyte-like cells

GMT

GATA4, MEF2C, and TBX5

NT-ESCs

nuclear transfer-embryonic stem cells

10.4161/org.28816