Prevention of tumor risk associated with the reprogramming of human pluripotent stem cells

ESCs are established from the inner cell mass (ICM) of blastocyst and can differentiate into all types of cells [15, 16]. The ability to produce teratomas in immune-deficient animals is common pluripotent properties of iPSCs and ESCs [1, 3, 37]. Tumors comprise different types of cancer cells, and this contributes to the heterogeneity of tumors [20]. Teratomas are defined as mixed benign tumors that comprise abnormally developed tissues derived from germ cells with normal karyotypes [37]. Teratomas are regarded as posing no direct danger of forming a malignant tumor, although they have the potential to metastasize in response to some interactions with their microenvironment and niches [38]. Substantial number of tumors can be generated by a series of mutations, which can cause uncontrolled cell division. The process of tumorigenesis is specified by alterations in genetic, epigenetic, cellular, and microenvironmental circumstances [20]. Therefore, the risk of tumorigenesis by any stem cell-derived transplants needs to be eliminated before their clinical applications.

To understand difference between and similarities in marker genes in CSCs and ESCs, we have summarized the corresponding signals and their characteristics (Table 1) [16, 17, 39].

A few examples are described below.

The uncontrolled proliferation of many tumors is driven by a small population of cancer cells, known as CSCs, which exhibit the capacity for self-renewal and pluripotency. Unlike somatic cancer cells, CSCs can produce an obvious cancer and propagate the malignant cancerous clones indefinitely. Like the stemness-related transcription factors, surface markers that are expressed in stem cells are also expressed in human cancers. These include TRA-1-60, Stage specific embryonic antigen-1(SSEA-1), Epithelial cell adhesion molecule (EpCAM), Aldehyde dehydrogenase 1 family, member A1 (ALDH1A1), Leucine-rich repeat-containing G-protein coupled receptor 5 (Lgr5), CD13, CD19, CD20, CD24, CD27, CD34, CD44, CD45, CD47, CD49f, CD66c, CD90, CD166, TNFRSF16, CD105, CD133, c-Kit, CD138, CD151, and CD166. Table 1 shows the surface markers of CSCs, some of which are targets of therapeutics in cancer treatment.

Three processes such as maintenance, self-renewal, and differentiation are concerned with embryonic development and homeostasis of adult tissues. Cancers commonly exhibit aberrant functions within these pathways, often in a cell context-dependent manner. Here we discuss the current evidence for the control of the Hedghog (Hh), Notch, LIF-JAK-STAT, PI3K-Akt-mammalian target of rapamycin (mTOR) and Wnt/β-Catenin pathways in CSCs.

The first event that triggers the transformation of normal cells into abnormal cells is mutation of the genome. Such mutations may be maintained, and other events, such as changes in the epigenome of stemness-related genes, oncogenes, and tumor suppressor genes, can also trigger the transformation into abnormal cells.

In general, “driver” mutations occur during the initiation stage of cancer and are followed by the accumulation of “passenger” mutations and tumorigenesis [245, 246]. Driver mutations trigger the growth and development of cancers, whereas passenger mutations do not affect clonality [247‐249]. Recent progress in the technology of deep sequencing has led to identify these mutations in particular oncogenes and/or tumor suppressor genes [250]. However, whether these mutations can become a barrier to the reprogramming of cancer cells remains unclear. Moreover, the reprogramming of cancer cells may induce genomic changes, including chromosomal aberrations, copy number variations (CNVs), and single-nucleotide variations. For example, chromosome abbreviations at trisomy 12, chromosome 8, and chromosome X have been found in ESCs and iPSCs [251‐256]. CNVs have been detected during reprogramming or mosaicism from parental cells. In some cases, CNVs are also lost by passaging of cells [254, 257‐260].

Single-nucleotide mutations have been analyzed using high-throughput next-generation sequencing technologies [261, 262]. Additional investigation is required to characterize the occurrence of these mutations during cell reprogramming. Mutations in mitochondrial DNA in human iPSCs increase with age, and the use of young donor cells may be one option to overcome this problem [263, 264].

The oncogenic potential of reprogrammed stem cells correlates with epigenetic and genomic instability [265, 266]. Epigenetic instability during cancer progression can lead to commitment to altered gene expression. Tumor generation can be caused by epigenetic reprogramming such as carcinogenic enhancer reactivation in both somatic and cancer cells [267‐269]. In general, DNA methylation can silence the expression of the tumor suppressor genes that are required for sustaining normal function, whereas aberrant expression of oncogenes can lead to cancer initiation [270‐272]. Intriguingly, iPSCs generated from human sarcoma cell lines have the same methylation and demethylation status in the promoters of oncogenes and tumor suppressor genes in the initial stages. In some cases, iPSCs can suppress the oncogenic promoters and maintain the activity of the promoter of tumor suppressor genes. These finding suggest that stem cell factors can inhibit the expression of cancer phenotypes, perturb epigenetics, and change cancer-related gene expression [266].

The factors that regulate chromatin are another possible target for circumventing the cancer risk during cellular reprogramming. Histone modification, noncoding RNA alteration, and chromatin alteration around mutated regions of DNA are key events that change oncogenic features [273]. Moreover, recent evidence of super-enhancers, locus control regions, and phase separation has provided new targets for next-generation reprogramming, to limit the oncogenic risk associated with cell reprogramming [273‐275]. The methodologies chosen to induce these epigenetic and chromatin changes require careful consideration.

The differences between CSCs and adult stem cells in the early stage are reflected in differences in stemness signals, such as BMP, notch receptor 1 (Notch 1), sonic hedgehog (Shh) signaling molecule, TGFβ, and the wingless-type MMTV integration site family (Wnt). Later, EMT factors such as HIFs, the zinc-finger protein SNAI2 (Slug), the zinc finger protein SNAI1 (Snail), the class A basic helix–loop–helix protein 38 (Twist 1), and the zinc-finger E-box-binding homeobox 1/2 (Zeb 1/2) are activated, which leads to changes in epigenetic signatures during progression. Changes in the epigenetic machinery might be crucial for the acquisition of stemness characteristics by, and the tumorigenesis of CSCs [276].

The epigenetic barriers that determine cell plasticity must be overcome during the initial steps of cell reprogramming. In the first step, DNA methylation by DNA methyltransferases (DNMTs) occurs at CpG islands and is reversed by ten-eleven translocation proteins (TETs), followed by transcriptional silencing. Both DNMTs and TETs affect transcriptional initiation, elongation, splicing, and stability at the CpG-poor and repeat-rich intergenic loci of target genes [277]. Some methylated DNA-specific binding proteins also perform similar functions. In the second step, histone modifications, such as methylation, acetylation, ubiquitination, phosphorylation, and SUMOylation, cause changes in nucleosomes. These changes define functional regions such as promoters, enhancers, and insulators, which determine the transcriptional patterns and cell fates. In the third step, histone modifiers are used to control chromatin during transcription, replication, and genome maintenance.

Several histone modifiers can regulate transcription, such as the polycomb repressive complex (PRC) 1 and 2, and the enhancer of zeste homolog 2 (EZH2), which is the catalytic subunit of PRC2 that mediates transcriptional repression by introducing H3K27me3 [278]. By contrast, in glioblastomas, the H3K27M mutation in H3.1 and H3.3 leads to reduced EZH2 activity and a decreased level of H3K27me3, followed by a more primitive stem-like state [279, 280]. Loss of EZH2 function can induce a transcription program for self-renewal and leukemogenesis [281]. Therefore, EZH2 mutation and deregulation of H3K27me3 seem to be linked.

The trithorax group (TrxG) complex is another factor that plays a role in the control of histone methyltransferase (HMT) in mixed-lineage leukemia (MLL) [282]. In this type of leukemia, the MLL fusion protein MLL-AF9, which lacks the catalytic domain, is produced by chromosomal rearrangement. The committed cells can be reprogrammed toward leukemic stem cells and initiate tumorigenesis [283]. MLL oncogenic fusion proteins require the repressive activity of PRC1, which mono-ubiquitinates histone H2A at lysine 119 (H2AK119Ubi) and mediates transcriptional repression in association with PRC2 [283]. BMI-1 in PRC1 is required to abolish tumor suppressor functions and to enhance CSC self-renewal in solid tumors [284]. Finally, ATP-dependent chromatin modeling complexes that move, eject, or restructure chromatin, are also key elements in the control of tumorigenesis.

Four subfamilies are involved in chromatin remodeling in tumor cells, such as switch/sucrose nonfermentable (SWI/SNF), imitation switch, chromodomain helicase DNA binding protein 1, and INO80 complex ATPase subunit. These differ in their functions, protein domains, and subunit constituents [285]. Loss of SWI/SNF-related matrix-associated actin-dependent regulator of chromatin subfamily B member 1, a subunit component of the SWI/SNF complex, results in genetic changes that drive rhabdoid tumors and are associated with the inhibition of differentiation, which leads to reprogramming toward oncogenic transcription for oncogenic signaling [286, 287]. The AT-rich interactive domain-containing protein 1A, another subunit of the SWI/SNF complex, functions as a tumor suppressor in colon cancers, and its deletion causes activation of the oncogenic transcriptional program for the promotion of invasive colon adenocarcinoma [288]. These data indicate that epigenetic modifiers also play a critical role in determining stemness/pluripotency and tumorigenesis.

Reversible epigenetic changes play a critical role in the fate decision in cancer cells, which can favor or disfavor the stem cell program that sustains tumor progression. Similarly, cell reprogramming involving epigenesis in cancer cells can generate iPSCs. Chromatin regulators such as the complex of PGC and TrxG proteins can regulate cancers and reprogramming [289]. In glioblastomas, several PRC and TrxG components play important roles in CSC development and in the plasticity of cancer cells. In adult glioblastoma, overexpressed MLL5 represses the expression of H3.3 in CSCs, which causes local reorganization of chromosomes [290]. The upregulation of MLL5 in non-stem cancer cells induces cell plasticity by inhibiting pro-neural differentiation, thereby eliciting a stem cell-like stage.

Another chromatin modifier, the linker histone variant H1.0, is also critical for the plasticity of cancer cells. Perturbation of H1.0 levels affects self-renewal activity directly and promotes the differentiation of non-stem cancer cells, thus blocking their tumorigenicity in vivo [291]. The plasticity of CSCs can be counteracted by epigenetic barriers that prevent cell reprogramming in vitro, such as the deposition of Suv39h1-associated H3K9me2/3 modifications [292]. An interesting insight into these epigenetic barriers was provided by a comparison of the epigenetics of gene expression between adult cells and CSCs in response to tissue damage [293]. In that report, tissue damage activated the resident stem cells, and CSCs exhibited overactivated stress-dependent enhancers and epigenetic modifications, which altered their cell fate and plasticity. These findings highlight the critical role of the reversibility of changes in the chromatin structure in the determination of the functional properties of CSCs.

The tumor microenvironment favors a stress-responsive enhancer, which may induce CSC plasticity. Therefore, the microenvironment and niche seem to be critical for the reprogramming to CSCs and iPSCs. Inflammation is an immune reaction to a pathogen [294], during which the responses of immune cells can lead to oxidative damage of DNA in infected cells. Nuclear factor kB and STAT3 can cause parenchymal cells to produce excess amounts of ROS and reactive nitrogen species, which can induce genomic instability and DNA mutations [295]. Mutations and chromosomal alterations are thought to be associated with tumor progression, which may be potentiated by a chronic inflammatory microenvironment that are damaged by mutations.

During CSCs have developed, CSCs may create their own niche. Cells forming the CSC niche necessary for both the maintenance of CSCs and the generation of factors and tumor-associated cells that maintain CSC properties such as invasion, metastasis, and promotion of angiogenesis [296]. The CSCs niche contains cellular components such as cancer-associated fibroblasts [297], tumor-associated macrophages [298], tumor-associated neutrophils [299], mesenchymal-associated cells [300], and cell-mediated adhesion and soluble factors [301], which played important roles in cell–cell communication.

Tumor suppressor genes are mostly transcription factors that modulate the antiproliferation signals that arrest the consistency of the cell cycle for DNA repair and that prevent mutation during cell division [302‐306]. For example, the p53 tumor suppressor gene protein product functions as a transcription factor and a cytoplasmic regulator in cell cycle arrest and apoptosis [306‐308]. The dysregulation of genes that modulate the cell cycle results in uncontrolled cell division, during which a series of mutations to proto-oncogenes and tumor suppressors are needed before cells transform to cancerous cells [309]. Proto-oncogenes are quiescent counterparts of oncogenes that become oncogenes upon mutation. The modified cells produce more of the gene product and exhibit excessive proliferative ability. Hormones or signal transduction can stimulate oncogenes to promote uncontrolled cell proliferation by changing the regulation of gene transcription [310, 311].

Pluripotent reprogramming factors are overexpressed in various cancers [312, 313]. OCT4 has been reported to be overexpressed in a gastric cell line [18], ovarian carcinoma, pancreatic cancer [18], prostate cancer [314], and bladder cancer [315]. c-MYC is also overexpressed in various cancers [316, 317] and can block differentiation and induce tumor formation in the absence of p53 [318, 319]. These observations indicate that reprogramming factors can act as potential proto-oncogenes in the reprogramming process and emphasize the potential risk of carcinogenesis by stem cell-like cells.

Reprogramming protocols have been used to inhibit the tumorigenesis induced during reprogramming in studies of various cancer cells [20‐22]. In these models, one or multiple sets of reprogramming factors are delivered to cancer cells and induced patient specific CSCs are generated. These models can be used to study therapeutic targets in the initial stage of carcinogenesis [20‐22]. In addition to the genetic methods that are used to generate pluripotent stem cells, some small molecules have been reported to promote cell reprogramming. For example, targeting signaling pathways like the HDAC, Wnt, and TGFβ cascades can regulate the expression of pluripotent stemness genes induce the reprogramming of cells [25]. Various molecules that promote reprogramming can replace genetic methods; these include recombinant reprogramming factors (OSKM) with polyarginine tags [26] and other small molecules [30‐33]. However, chemically defined small molecules that induce reprogramming alone have not been developed for human induced pluripotent stem cells (hiPSCs) [24]. Furthermore, whether small molecule-driven iPSCs can prevent the risk of tumorigenesis when used therapeutically has not been determined. The ability of undifferentiated iPSCs to produce teratomas in grafted cell populations is one of the main concerns of this approach, and the nature of these cells should be clarified genetically. Precise information about these transplantation animal models, inoculated patient-specific iPSCs, and the resultant organoids is needed.

The use of clones that have been pre-evaluated as being safe is one possibility for overcoming carcinogenic activity, as demonstrated in neural stem cells (NSCs) [320‐322]. An antibody against SSEA-5 on human ESCs (hESCs) can be utilized to remove teratoma-forming cells from dissociated hESCs [323]. In immune-deficient mice, flow cytometric analysis has shown that transplants of SSEA-5-negative cells formed smaller teratomas, measuring < 1 cm³, whereas SSEA-5-positive transplanted cells formed teratomas > 1 cm³. Flow cytometric analysis using an anti-SSEA-5 antibody may be useful for separating undifferentiated, unsafe cells from iPSC-driven mixed-cell populations. The induction of senescence generated by replication stress and DNA damage, telomere shortening, environmental and oncogenic stresses, and the pro- inflammation inducing microenvironments is one of the oncogene-associated events involving epigenetic control [285]. Cancer-associated senescence has been reported to promote cancer stemness and plasticity in CSCs [306].

Three different approaches for the removal of tumor-initiating pluripotent stem cells from ESCs and their differentiated cells have been reported: (i) chemical treatment [324‐327], (ii) genetic treatment [328‐330], and (iii) immunological treatment [323, 331‐335] (Table 2). Each method has its advantages and disadvantages; the latter include the high cost, variation between lots, nonspecific antibody binding, integration of toxic genes, and long duration [354]. A strategy for application to clinics to abate the teratoma formation should be explored further [354, 355]. The use of small molecules has several advantages, such as robustness, efficiency, speed, simplicity, and low cost. Some small molecules can inhibit the formation of teratoma in human pluripotent stem cells (hPSCs). For example, an inhibitor of stearoyl-CoA desaturase (SCD), PluriSIns #1, has been shown to prevent teratoma formation [325]. SCD1 is an important enzyme in the biochemical synthesis of monosaturated fatty acids, which are needed for the survival of hPSCs. A recent study reported that SCDs play a critical role in endoplasmic reticulum (ER) stress and promote the survival of glioblastoma cancer stem cells [356]. The N-benzylnonanamide JC101 induces ER stress via the protein kinase RNA-like endoplasmic reticulum kinase/ATF4/DNA damage inducible transcript 3 (DDIT3 = CHOP) pathway [340], which leads to the inhibition of teratoma formation. In a study that used iPSCs to treat spinal cord injury, these cells were determined to be safe before the grafting of transplants, which prevented the formation of teratomas [357].

The introduction of inhibitors of antiapoptotic factors that effectively remove residual iPSCs can also prevent teratoma formation [344, 348]. Treatment with survivin (an antiapoptotic factor) and a novel survivin suppressant, YM155, was effective in decreasing the risk of teratoma formation. In that study, addition of YM155 permitted the survival of CD34⁺ cord blood cells and prevented teratoma formation in human induced pluripotent stem cells (hiPSCs)-grafted mice [347].

The introduction of inducible caspase-9 (iCap9), as a suicide gene, into hiPSCs avoided tumorigenic transformation after their transplantation [347, 348]. The efficiency of iCap9 and small molecule-like chemically induced dimerization (CID) in the prevention of the risk of tumorigenesis was evaluated in cell-transplantation experiments. iCap9 integrated with CID induced the apoptosis of iPSCs and iPSC-derived NSCs and produced the terminal differentiation of transduced cells grafted into injured spinal cord in mice but avoided the formation of teratomas [348].

A conditionally replicating adenovirus that targets cancers using multiple factors (m-CRAs) is a new antitumorigenic agent used in hPSC-based cell therapy. Given that the survivin promoter is stronger in undifferentiated hPSCs than was the telomerase reverse transcriptase (TERT) promoter, surviving promoter m-CRAs efficiently kill undifferentiated hPSCs but not differentiated normal cells when compared with TERT promoter m-CRAs. The surviving promoter m-CRAs seems to be a novel antitumorigenic agent that may facilitate safer hPSC-based medical applications [358].

Another approach is the isolation of the desired differentiated cells from other cell types and undifferentiated hPSCs, such as the removal method of the residual pluripotent cells from other cells using fluorescence-activated cell sorting or antibodies coated magnetic beads against a particular antigen, including SSEA-5 [323], claudin-6 [333, 359], and the Ulex europaeus agglutinin-1 fucose-binding lectin (UEA) [360]. Caludin-6 bound to the enterotoxin produced by Clostridium perfringens kills the hPSCs that induce teratoma formation [333]. Yet another approach is the direct, targeted killing of tumor cells using a cytotoxic antibody against the podocalyxin-like protein-1, an inhibitor of stearoyl-CoA desaturases or a DNA topoisomerase II inhibitor, etc. Etoposide, a DNA topoisomerase II inhibitor, and the CDK inhibitor purvalanol have been used to minimize the risks of tumor formation by post-transplanted ESCs and iPSCs [345, 346]. Lin et al. reported that the cardiac glycosides digoxin and lanatoside C can kill undifferentiated hESCs [339].

The activation of the Wnt signaling is critical for maintaining of pluripotency in ESCs [41, 185, 186, 192, 193]. JDP2, has been used as a pluripotency-promoting factor in the reprogramming of cancer cells to reduce their cancerous features. JDP2 can activate or repress various gene transcription actions [185, 361], although it also plays a critical role in malignant transformation. For example, JDP2 can promote cancer cell growth in leukemia and hepatocellular carcinoma [362, 363] but may also be a tumor suppressor in cancer cells [159, 364]. The treatment of gastric cancer cell lines with OCT4 and JDP2 inhibited the tumorigenic function of the cells by switching off BMP7 [195, 365]. The tumor-forming ability of these partially reprogrammed gastric CSC-like cells is reduced in immune-deficient mice, which shows that JDP2 plays a critical role in the reduction of oncogenic potential [185, 189].

Double deficiency of ATF3 and Jdp2 in mice stromal tumors promotes cancer growth [361]. Jdp2 is regarded as a factor that drives reprogramming through the regulation of the Wnt signaling pathway and the suppression of ROS production [188]. Oxygen regulates pluripotent stem cells via Wnt–β-catenin signaling and HIF-1α [188]. Wnt signaling plays a key role in the generation and maintenance of iPSCs [366]. JDP2 is one of the target genes of the LEF–TCF family and is involved in the Wnt signaling pathway [185]. Wang et al. generated iPSC-like cells by introducing OCT4 and JDP2 (JO) into gastric cancers for reprogramming [306]. ESC-like colonies were obtained in teratomas of JO-introduced xenografts. The JO-introduced xenografts were smaller and had fewer necrotic and mitotic cells, and a larger nucleus, than the parent cancer xenografts.

The roles of EZH2, HMT, and MLL in the regulation of BMP7 expression have been examined [364, 365]. The findings of this study suggest that HMT inhibitors can be used to prevent MLL and that EZH2 is a potential therapeutic target. The β-catenin–JDP2–RPMT5 complex axis is critical for reestablishing glutathione homeostasis after genotoxic-mediated stress and may be useful in the development of drugs aimed at treating or avoiding inducible resistance to chemotherapy in cancer cells [367]. A combination of factors (Jdp2, Jhdm1b, Mkk6, Glis1, Nanog, Essrb, and Sall4) was reported to be a useful reprogramming complex for efficiently leading mouse embryonic fibroblasts (MEFs) to become chimera-competent iPSCs in xenografts [368]. These findings suggest that the direct reprogramming of stem cells cannot reduce the risk of malignancy itself, and that excluding cancer cells from the microenvironment of stem cells may be crucial for reducing the risk of cancer.

Methods of pluripotency reprogramming that do not use genetic material provide another potential strategy for generating safe iPSCs. To date, various molecules that promote cell reprogramming have been reported as substitutes for genetic materials. These small molecules modulate the activities of the enzymes that are involved in the epigenetic modification of genes and may play a crucial role in pluripotency reprogramming [25, 27‐30, 369]. These small molecules include the TGFβ inhibitory antagonist SB431542, the MAPK–extracellular signal-regulated kinase inhibitor PD0325901, and the inhibitor of Rho-associated coiled-coil-containing protein kinase thiazovivin (2,4-disubstituted thiazole or TZV). TZV significantly increases the rate of hiPSC reprogramming by up to 200-fold [370]. The combination of Oct4, the TGFβ receptor inhibitor A83–01, and the methyltransferase inhibitor AMI-5 increases the reprogramming efficiency of mouse iPSCs, and the repression of ROS increases the efficiency of cell reprogramming [371, 372]. The addition of vitamin C to the culture medium significantly increases the reprogramming efficiency. However, the generation of hiPSCs using small-molecule compounds alone has not been shown to reduce the risk of tumorigenesis significantly. Further investigation of nongenetic compound-guided iPSCs is required to exclude the potential for tumor progression before patient engraftment.

Vitamin C is a candidate for preventing the tumorigenesis caused by hiPSCs because it inhibits the improper self-renewal of human stem cells by increasing histone demethylase activity. Vitanin C has also been applied as an anticancer agent in melanomas [373‐376]. The histone demethylases JIhdm1a/1b are the direct downstream molecules of vitamin C. In addition to its antioxidant activity [376], vitamin C was reported to increase the rate of reprogramming of fibroblasts to iPSCs when added to the culture medium [190]. In that study, supplementation with vitamin C decreased p53/p21 levels, which successfully blocked pluripotency reprogramming.

The mutation frequencies in hiPSCs increases with age. Aging is also one of the major risk factors for cancer progression. Therefore, younger donor cells may be used to avoid detrimental mutations in mitochondrial DNA. The reprogramming of somatic cells into iPSCs resets the cellular damage and the stress- and senescence-associated epigenetic marks in vitro, which suggests that the reprogramming technology may be able to induce the rejuvenation of senescent cells [377]. Mosteiro et al. reported that cells reprogrammed in vitro were present next to senescent cells clustered in reprogrammable mice [378]. These senescent cells promote reprogramming through senescence-associated secretory phenotypes, especially via the production of interleukin 6 [379]. The functional crosstalk between reprogramming and senescence has been reported elsewhere [380, 381].

To avoid the risk of point mutations in stem cells, histocompatibility antigen-matched umbilical cord blood-derived iPSCs are useful for ensuring a minimal rate of mutations [382]. Amniotic cells derived from human placental tissues exhibit lower immunogenicity and anti-inflammatory properties than do other cells [383, 384]. The expression of putative immunosuppressors, such as CD59 and CD73, is repressed during the process of reprogramming in amniotic cells [385]. These findings emphasize that long-term-cultured hESCs and hiPSCs show increased growth rates and decreased dependence on the concentration of growth factors and malignant tumor-forming abilities after their engraftment in mice [386]. Carcinoma-derived iPCSCs, which have a higher number of passages, exhibit increased carcinogenic ability compared with those obtained from donor cholangiocellular carcinoma [387].

Other candidate for therapeutic agents includes drugs against the super-enhancers involved in phase separation [388]. In general, the machinery used for epigenetic modification can be altered during transcription initiation, elongation, and termination by involving the transcriptional complex containing polymerase II. Transcriptional regulators such as the mediator MED, coactivator Brd4, splicing factor SRSF2, histone heterochromatin protein 1, and nucleoplasmin 1 may be targets for therapies aimed at modulating chromatin structure during reprogramming and for blocking the development of cancer or other genetic diseases [389]. Hepatocellular carcinomas (HCCs) exhibit distinct promoter hypermethylation patterns. Deletion of C/EBPβ expression using the CRISPR-Cas9 system affects the co-recruitment of BRD4 at enhancers marked with H3K27ac to block the development of HCC [390].

The two mediator kinases, CDK8 and CDK19, may enable transcriptional reprogramming through the super-enhancer family, chromatin looping by CTCF, and cohesion with the transcriptional mediators MED12 and 13, which may be potential drug targets [391]. The valproic acid-mediated inhibition of HDAC and glycogen synthase kinase-3 results in the transcriptional repression of many genes, including the gene that encodes for myristoylated alanine-rich C-kinase substrate 1, which is the Actin-stabilizing protein that is required for the process of the early development of dendritic morphogenesis and synapse maturation [392]. In zebrafish, p300 and Brd4 trigger genome-wide transcriptional network by controlling histone acetylation of the first zygotic genes. This mechanism is crucial for initiation of zygotic development and reproduction [393]. In mammals, these target genes are also potential candidates for drug targets. Stem cell-derived organoids may also be useful for identifying the effects of the microenvironmental niche on cancer development and the reprogramming fates of stem cells [394].

The successful epigenetic reprogramming of primary cancerous cells was recently reported [350]. Pancreatic ductal adenocarcinoma cells were reprogrammed by independent protocols, such as the introduction of OCT4 together with miRNA of Mir302, the episomal vector-encoded NANOG and REX1, in Yamanaka’s protocol (as a control). These methods yielded efficient reprogramming with reduced tumorigenicity through epigenetic changes, such as downregulation of TET2, SIRT1, disruptor of telomeric silencing 1-like (DOT1L), and T-box 3 (TBX3) and upregulation of TET1. A study of ischemic cardiomyopathy found that DNA methylation and the metabolic pathway were strongly correlated through the EZH2–KLF15 axis [351]. C/EBPβ recruits the DOT1L methyltransferase to open chromatin via methylation of H3K79 in multiple drug-resistance genes [352].

Imprint dysregulation compromises the developmental ability of pluripotent stem cells [353]. The stability of de novo methylation of CpG islands in PSCs is critical for cancer development. Thus, CpG islands and imprinting-control regions are important for the evaluation of development, stemness, and pluripotency [395]. Reprogrammed hiPSCs exhibit greater loss of imprinting compared with ESCs, and the loss of imprinting preexists in their somatic cells of origin. The differently imprinted genes related to the loss of imprinting, especially those of paternal control, are more prone to disruption [396].

Ubiquitin-like with PHD and RING finger domain 1 (Uhrf1) is a hemimethylated DNA-binding protein that interacts with DNMT1 and recruits euchromatic histone–lysine N-methyltransferase 2 to form heterochromatin, together with tripartite motif-containing 28 and HDACs, for DNA methylation. In mouse stem cells, Uhrf1 interacts with the SET domain complex containing 1A/COMPASS, followed by positive regulation of H3K4me3 modification. Uhrf1 maintains bivalent histone marks, such as H3K27me3 and H3K4me3, and particularly those associated with specifications to the neuroectoderm and mesoderm [396]. Tatton-Brown–Rahman syndrome (TBRS) is caused by a mutation in DNMT3A, a gene that is also associated with Sotos syndrome, which is caused by haploinsufficiency of NSD1, an HMT that catalyzes the demethylation of histone H3 at K36 (H3K36m22). In the mouse, NSD1-mediated H3K36me2 methylation is induced by the recruitment of DNMT3A and is crucial for the maintenance of DNA methylation at intergenic regions. The binding of DNMT3A to H3K36me2 can be inhibited by a missense mutation associated with TBRS, which suggests that trans-chromatin regulatory control is critical for human neoplastic and developmental growth [397].