SEARCH

SEARCH BY CITATION

Keywords:

  • MicroRNA;
  • Reprogramming;
  • Induced pluripotent stem cells;
  • Transdifferentiation

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Induced Pluripotency
  5. Conclusions
  6. Author Contributions
  7. Disclosure of Potential Conflicts of Interest
  8. References

It is now well-established that somatic cells can be reprogrammed to alternative cell fates by ectopic coexpression of defined factors. Reprogramming technology has uncovered a huge plasticity besides gene regulatory networks (GRNs) of differentiated cell states. MicroRNAs (miRNAs), which are an integral part of GRNs, have recently emerged as a powerful reprogramming toolbox. They regulate numerous genes, thereby modulating virtually all cellular processes, including somatic cell reprogramming. Not only can miRNAs provide novel opportunities for interrogating mechanisms of induced pluripotency and direct lineage reprogramming but they also offer hope for the efficient creation of safe cell sources for regenerative medicine. In reviewing landmark roles of miRNAs in cell reprogramming, we offer suggestions for evolution of the reprogramming field. Stem Cells 2014;32:3–15


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Induced Pluripotency
  5. Conclusions
  6. Author Contributions
  7. Disclosure of Potential Conflicts of Interest
  8. References

In vivo, genuine pluripotent cells are found in small numbers and are only ephemerally accessible, thus embryonic stem cells (ESCs) provide an interminable cell source to interrogate pluripotency and self-renewability in vitro. Induced pluripotent stem cells (iPSCs), which circumvent ethical and immunological problems associated with ESCs, offer an amazing, alternative tool to investigate the gene regulatory network (GRN) of pluripotency, discover new drugs, establish disease models “in a dish,” and provide a proper cell source for the future of regenerative medicine. As an alternative to iPSC technology, another promising cell source for personalized cell therapy could be from direct reprogramming of one developmentally restricted cell type into another without going through a pluripotent state [1]. This new strategy has the advantage of bypassing the potential tumorigenicity associated with pluripotent stem cell application. As shown in Figure 1, a variety of approaches and reprogramming molecules have been used to establish, in somatic cells, a new GRN that promotes pluripotency (pluripotent reprogramming) or other differentiated fates (direct lineage reprogramming).

image

Figure 1. Molecules and strategies in somatic cell reprogramming. To reprogram cell fates, master molecular players such as transcription factors (TFs) and miRNAs enriched in desired cell types are introduced into starting cells such as fibroblasts, keratinocytes, or blood cells via different viruses which efficiently deliver and highly express exogenous factors. Some viruses integrate the transgenes into the genome (e.g., retroviruses and lentiviruses) and/or endanger the safety of reprogramming because of their viral entity (all viruses including retro-, lenti-, adeno- and Sendai viruses). Integrative vectors can substitute for viruses, but they still cause genomic integrations. Excisable integrative vectors such as transposons and Cre-loxP vector systems and nonintegrative vectors that include episomal vectors and mini-circles provide alternative safer tools to deliver reprogramming factors. Modified mRNAs of the reprogramming factors have also been used to generate vector-free iPSCs, however they may be reverse transcribed and integrated back into the genome. Direct delivery of the reprogramming proteins provides another safe approach to induce fate switch, but their preparation is labor-intensive, resulting in an extremely low efficiency. In diverse reprogramming strategies, small-molecule chemicals are frequently used to either augment reprogramming or replace one or more of the reprogramming factors. Small molecules easily enter the cells and, if do not cause genotoxicity, would provide a safe means to promote cell state transition. miRNAs are approximately 22-nucleotide RNAs that control hundreds of genes and virtually all cellular pathways. They can be misexpressed by various viruses and vectors. Additionally, miRNA precursors or mature miRNA mimics can be transiently transfected into cells. Importantly, due to their small size, mature miRNAs cannot be reverse transcribed for integration into the genome. In conjunction with other safe approaches such as small molecules, recombinant proteins, or episomally encoded TFs, miRNAs may yield the best results for safe reprogramming into pluripotency and other cell states. Abbreviations: iPSC, induced pluripotent stem cells; mRNA, messenger RNA; miRNA, micro RNA.

Download figure to PowerPoint

Reprogramming technology has disproven the belief that established cell fates are irreversible, uncovering an immense plasticity besides robust GRNs of differentiated cell states. At the molecular level, cell fate plasticity is orchestrated by integrated GRNs that govern cells' behavior and developmental identity. The core GRNs that control the developmental fates of diverse cells are composed of specific sets of transcription factors (TFs), epigenetic regulators, signaling mediators, and post-transcriptional regulators that include noncoding regulatory RNAs. These GRN components dynamically regulate each other in different regulatory loops to confer robustness to cell fate decisions [2-5]. Thus, rewiring/reprogramming of cellular GRNs is expected to render different cell fates interchangeable. Recently, a class of noncoding RNAs known as microRNAs (miRNAs) has emerged as critical regulators of the GRN architecture [6, 7] and novel potent players in the reprogramming arena. miRNAs are approximately 22-nucleotide regulatory RNAs that, through Watson-Crick base-pairing, interfere with translation of their target mRNAs into proteins by directly repressing translation or inducing mRNA destabilization/cleavage [6] (Fig. 2A). They are predicted to regulate over half of the mammalian transcripts [8], thereby influencing virtually all aspects of cellular life, including fate specification and determination. miRNAs are dynamically regulated and accomplish critical functions during cell fate transition. Cumulative evidence indicates that modulating miRNAs, along with or without specific TFs, could promote cell fate switch in somatic cells toward pluripotency or other alternative cell states. In this review, we describe our current knowledge on the roles of miRNAs during iPSC generation and lineage transdifferentiation. We rationalize how these intriguing RNAs may provide both a deeper insight into mechanisms of cell fate reprogramming and a promising tool for the future of regenerative medicine. Finally, in reviewing essential and indispensable roles of miRNAs in cellular reprogramming, we offer suggestions for the evolution of the reprogramming field.

image

Figure 2. miRNA behavior in iPSC generation. (A): miRNA biogenesis. miRNA genes are located either in intergenic or intragenic regions of the genome and are transcribed mostly by RNA polymerase II. More than 50% of miRNAs lie within the introns of protein-coding genes (so-called host genes). Many of the miRNA genes are organized in tandem as clusters, being transcribed as polycistronic miRNA transcripts (e.g., miR-302–367 cluster). On the other hand, other miRNAs are individually encoded by single genes (e.g., miR-34a). miRNAs form a post-transcriptional mechanism of gene regulation by which target mRNAs are either degraded or translationally repressed. Following incorporation into the so-called miRNA-induced silencing complex (miRISC), they find and bind their numerous mRNA targets through imperfect complementarity to specific binding sites. The miRNA seed sequence (5[prime] nucleotides 2–8) is the main determinant of miRNA target finding. The degree of base pairing between a miRNA and its bound target mRNA determines whether the mRNA will be degraded or translationally inhibited. (B): Patterns of miRNA expression during iPSC generation. During reprogramming to pluripotency, pluripotency-associated miRNAs are upregulated, while those associated with differentiation are downregulated. Expectedly, the targets of pluripotency miRNAs decrease and those of differentiation miRNAs increase over the course of iPSC induction. Abbreviations: iPSC, induced pluripotent stem cell; miRNA, micro RNA; RISC, RNA-induced silencing complex.

Download figure to PowerPoint

Induced Pluripotency

  1. Top of page
  2. Abstract
  3. Introduction
  4. Induced Pluripotency
  5. Conclusions
  6. Author Contributions
  7. Disclosure of Potential Conflicts of Interest
  8. References

miRNAs: Dynamic Expression and Critical Functions

Through a groundbreaking work, Takahashi and Yamanaka demonstrated that delivery of a few pluripotency factors (Oct4, Sox2, Klf4, and c-Myc; OSKM) is sufficient to induce pluripotency in differentiated cells [9]. Despite heady progress, the mechanism of induced pluripotency has not been completely understood. Following delivery, reprogramming pluripotency factors are immediately confronted by a stable differentiated GRN, reorganization of which being necessary to establish a new GRN and induce pluripotency. To accomplish this, the reprogramming factors engage with chromatin, interfere with the parental cell's molecular complexes, regulate their target genes, including miRNAs, and gradually impose their control over the parental cell GRN. iPSCs express a specific set of miRNAs similar to that of ESCs, yet distinct from that of starting cells, and are clustered with ESCs in “miRNAome” [10-12]. miRNAs exhibit a dynamic, biphasic pattern of expression during iPSC production [13], such that two transcriptional waves occur for miRNA genes [12, 13]. Interestingly, miRNA downregulation in both waves was more remarkable than miRNA upregulation [13], probably suggesting that silencing of the somatic program either precedes or is as important as activation of the pluripotency program in induced pluripotency. Unlike reprogramming-competent cells, reprogramming-refractory cells failed to activate the second transcriptional wave [13].

In reprogramming-competent cells, some miRNAs are gradually upregulated (e.g., miR-182) or downregulated (e.g., miR-34c and miR-214) [12-14]. Other miRNAs are upregulated or downregulated immediately after OSKM delivery, yet exhibit a late decline (e.g., miR-135b, miR-106a, miR-29a/b, miR-181) or rise (e.g., miR-142-3p, miR-494) in their subsequent mode of expression [12-15] (Fig. 2B). Functionally, gain of miR-181, miR-182, miR-106a, and miR-29b functions [13-17] as well as loss of miR-34c and miR-29a functions [18, 19] were found to increase reprogramming efficiency. In contrast, gain of miR-214 function decreased reprogramming efficiency [13]. Overall, miRNAs associated with pluripotency (e.g., miR-17 family, miR-290–295 cluster, miR-302–367 cluster) are upregulated [13, 16, 17] and those associated with differentiation (e.g., let-7 and miR-34 families) are downregulated during reprogramming [13, 18]. As might be expected, both predicted and validated targets of differentiation-associated miRNAs increase and those of pluripotency-associated miRNAs decrease over the course of reprogramming [13]. Notably, inhibition of differentiation-associated miRNAs as well as ectopic expression of pluripotency-associated miRNAs during iPSC formation improves reprogramming efficiency [18, 20-22].

In starting somatic cells, OSKM attenuates somatic cell-enriched miRNAs, yet gradually induces the expression of miRNAs that are enriched in ESCs [12, 13, 19]. Cells in the early phase of reprogramming display a pattern of miRNA (and mRNA) expression that more closely resembles that of starting mouse embryonic fibroblast (MEF) cells and thus cluster more closely with MEFs [12, 23]. As reprogramming proceeds, the miRNA expression program transitions into a pattern that is (and progressively becomes) more similar to that of ESCs and bona fide iPSCs [12, 13]. miRNAs are indispensable for reprogramming to pluripotency. Indeed, miRNA deficiency during iPSC “induction” greatly decreases reprogramming efficiency [12, 16] or, when Dicer is knocked out, completely abolishes iPSC generation [24]. More importantly, knockdown of a single miRNA cluster, the miR-302–367 cluster, during reprogramming diminishes and its knockout “completely” blocks iPSC generation [25]. Taken together, miRNAs are differentially expressed and accomplish important functions during the pluripotent reprogramming of somatic cells.

miRNAs Modulate Reprogramming Barriers

The generation of iPSCs is hindered by numerous barriers, many of which are modulated by miRNAs. In principle, key molecular players enriched in starting cells as well as some of the genes inadvertently activated by reprogramming factors emerge as reprogramming barriers. At the initial stage of reprogramming, OSKM activates the Ink4/Arf and p53 loci, knockdown or knockout of which dramatically increases reprogramming efficiency [26-28]. Additionally, miRNAs inhibiting the p53 pathway (e.g., miR-138 and miR-17 family members) augment iPSC formation [16, 29]. p53 exerts its functions through its downstream targets including the miRNAs miR-199a-3p [30] and the miR-34 family members [18, 31]. The p53-regulated miRNAs are highly induced in early reprogramming cells, building big barriers to “efficient” somatic cell reprogramming by targeting critical pluripotency factors (Fig. 3). Of note, unlike p53 deficiency, the deficiency of p53-regulated miRNAs leads to stable, fully pluripotent iPSCs [18, 30]. This suggests the importance of p53 activity in generating safer, higher-quality iPSCs.

image

Figure 3. miRNAs and the p53 pathway. Following activation by OSKM, p53 and its downstream effectors antagonize efficient reprogramming to pluripotency by ensuring the integrity of the resultant iPSCs. p53 pathway exerts its reprogramming-inhibitory effects by inducing G1 arrest (through p21 and miR-199a-3p) and targeting key pluripotency factors such as Nanog and Sox2 (through the miR-34 family). Additionally, by targeting Klf4, the p53-p21 axis resists MET, a critical phenomenon in the early reprogramming phase. ESCC miRNAs dampen the G1-arresting effect of this pathway by blocking the G1 restriction checkpoint and establishing an ESC-like cell cycle program. Moreover, they promote MET by targeting key mesenchymal markers. Activations are indicated by arrows, inhibitions by blunted lines, and delayed inhibitions by dotted lines. Abbreviations: ESC, embryonic stem cell; ESCC, ESC-cycle-regulating miRNAs; iPSC, induced pluripotent stem cell; miRNA, microRNA; MET, mesenchymal-epithelial transition; OSKM, Oct4, Sox2, Klf4, and c-Myc.

Download figure to PowerPoint

Another reprogramming barrier is created by mesenchymal molecules, and a mesenchymal-to-epithelial transition (MET) has been observed to occur and to be necessary early in reprogramming [32, 33]. miRNAs have been found to contribute to this process, thereby critically modulating iPSC generation. The miR-200 family is directly induced by Oct4 and Sox2 early in reprogramming and aids in MET by targeting mesenchymal markers, including ZEB2, thereby derepressing epithelial markers such as E-cadherin [34]. In addition, miR-29b, which is directly induced by Sox2 in the early phase, augments MET, as a DNA methylation-associated phenomenon [32], through the direct repression of de novo DNA methyltransferases (Dnmts) [14].

OSKM has also been shown to confer additional robustness to this early-phase-associated MET by inducing the expression of miR-17 family [16, 17]. Several members of the miR-17 family as well as miR-302s belong to ESC-cycle-regulating (ESCC) miRNAs which share a common ESCC miRNA seed sequence and contribute to the unique cell cycle of ESCs by targeting critical cell cycle regulators [35]. Ectopic expression of ESCC miRNAs has been found to reinforce MET by suppressing mesenchymal markers (including TgfβrII) and upregulating epithelial markers (including E-cadherin) [16, 17, 21, 22]. Also, miR-467d with a slightly divergent ESCC miRNA seed sequence enhanced reprogramming efficiency [15]. Indeed, pluripotency-associated miRNAs increase reprogramming efficacy not only via ensuring MET and modulating the cell cycle but also by targeting several other direct targets such as NR2F2, MBDII, MeCP2, Wwp2, Fbx7, Pten, Cdkn1a, Nr2c2, Marcks, and LSD1 which provide barriers for successful reprogramming to pluripotency [15, 36-39].

The murine miR-290–295 cluster also harbors ESCC miRNA members but is not involved in MET during OSKM-based iPSC generation. In the early phase, its genomic region is epigenetically nonpermissive and thus remains silent. Instead, it is highly induced by c-Myc in the “late” phase of reprogramming. Moreover, ESCC members of this cluster could replace c-Myc in the transformation of somatic cells into a “homogenous” population of iPSCs [22]. Given that c-Myc is dispensable in alternative reprogramming cocktails [40, 41], other mechanisms might act redundantly to activate this cluster at the appropriate time.

The let-7 family of miRNAs is highly expressed in differentiated cells and oppose pluripotency by repressing key pluripotency factors such as c-Myc, n-Myc, and Sall4, the miR-290–295 cluster, and the RNA-binding protein Lin28 [20]. Although Lin28/miR-290–295 cluster and let-7 reciprocally antagonize each other, this regulatory paradigm does not exist in the early reprogramming phase, as Lin28 and miR-290–295 cluster are induced very late during iPSC induction [42]. By antagonizing let-7, Lin28 may provide one late mechanism (besides induction by c-Myc) through which the miR-290–295 cluster could be activated, which aids in establishing an ESC-like cell cycle program in culture. Notably, despite its low, if any, levels in the early reprogramming phase, Lin28 has been shown to promote reprogramming when combined with OSN (N: Nanog) [41]. Similarly, although the ESCC miR-302–367 cluster is induced only late in reprogramming, its ectopic expression in the early phase enhances reprogramming efficiency and kinetics [13, 15, 17, 22].

Several of the same miRNAs which cause, or are induced upon, ESC differentiation, resist iPSC formation as exemplified for miR-9, miR-34, let-7, and miR-199a-3p [18, 20, 30, 43, 44]. So, because hsa-miR-181, miR-134, miR-470, miR-145, and miR-296 are induced upon ESC differentiation and directly repress the pluripotency-associated TFs, NANOG, SOX2, OCT4, KLF4, SIRT1, LRH1, and CARM1 [45-48], they may be similarly implicated in the regulation of iPSC formation. On the other hand, ESC-enriched miRNAs facilitate reprogramming. The miR-302–367 cluster, hsa-miR-371–373 cluster, miR-205, mmu-miR-290–295 cluster, primate-specific hsa-miR-519a, miR-301/130/721 family, miR-200 family, and miR-17 family have all been found to highly increase reprogramming efficiency when ectopically expressed [16, 17, 21, 33, 34, 37, 49, 50]. Importantly, some of these miRNAs could replace the oncogene c-Myc [17, 22, 49], however, they were unable to significantly increase reprogramming efficiency in the presence of c-Myc [17, 22, 50]. This suggests that c-Myc may regulate those miRNAs. In agreement with this, c-Myc has been reported to stimulate specific miRNAs associated with pluripotency and tumorigenesis (e.g., miR-17 family) [16, 17, 51, 52]. Notably, knockdown or knockout of a single pluripotency-associated miRNA cluster, the miR-302–367 cluster, decreases reprogramming efficiency or completely abolishes iPSC production, respectively [25]. Of note, although mmu-miR-181 family is not an ESC-enriched miRNA family [4] and can compromise mouse ESCs upon ectopic expression [53], it is induced early in reprogramming and (its introduction into the somatic starting cells) promotes the early phase of reprogramming [15] (Fig. 2B). While mmu-miR-181 was found to increase the number of OSK-reprogrammed colonies with the same size as the control group, ESCC miRNAs (miR-294) increased both the efficiency and kinetics of reprogramming, yielding significantly larger colonies [15]. It is likely that the emergence of larger colonies by ESCC miRNAs is at least partly due to their targeting of the Rb family members which suppress ESC colony expansion by regulating the G1-S restriction point of the cell cycle [54]. Cointroduction of mmu-miR-181 and miR-294 showed no synergetic effects on iPSC production, but in spite of targeting independent sets of reprogramming barriers, they were found to converge on a subset of signaling pathways (activation of Wnt and Akt signaling axes and suppression of TGFβ pathway), thereby promoting the early stage of iPSC generation [15].

Although miRNAs potentially target numerous transcripts, only a few direct targets have been validated for the miRNAs that modulate induced pluripotency (Table 1). By controlling various targets, miRNAs synergize with pluripotency TFs in diverse regulatory loops to critically influence all important aspects of iPSC induction (Fig. 4). Identification of the targetome of miRNAs involved in reprogramming will provide a deeper understanding of the molecular basis of cell fate transitions and will help improve the methodological approaches to cell reprogramming. Importantly, because of their ESC specificity and regulating hundreds of genes, ESC-enriched (-specific) miRNAs have motivated stem cell researchers to investigate their reprogramming capability in the absence of exogenous TFs. The miR-302–367 cluster along with or without other miRNAs (miR-200c and miR-369s) has been reported to reprogram a number of mouse and human somatic cells to iPSCs in the absence of exogenous TFs [38, 55-57]. The safe, transient delivery of these miRNAs resulted in a fully reprogrammed pluripotent state, but with an extremely low efficiency (0.03%) and a high proportion of partially reprogrammed colonies (0.3%) [57]. The inefficiency of this approach might be attributable to the need for repetitive delivery of miRNA mimics into the cells [57], which could cause oscillations in gene regulation. Although the lentivirally encoded miR-302–367 cluster has been reported to efficiently promote somatic cell reprogramming alone [56], further optimizations to this approach appear necessary, as the attempts of three independent research groups to reproduce the reported results in somatic cells have recently met with failure [25, 36, 37]. Precise regulation of miR-302–367 cluster levels may also be important, as it has well-known targets among cell cycle regulators and induces cell cycle arrest and/or apoptosis if overexpressed ectopically [38, 55, 58]. Alternatively, the miRNA cassette construction and the delivery approach may affect both the miRNA processing and the efficiency of miRNA delivery into cells. Of note, this cluster, although not capable of inducing pluripotency alone, was potent enough to replace c-Myc in human iPSC generation [25], rescued partially reprogrammed iPSCs, and enhanced reprogramming efficiency quite remarkably [17, 36, 37]. Overall, the miRNA-only strategy of pluripotency induction deserves more attention and may be proven to be more applicable if optimized and combined with other iPSC-promoting factors/agents; this latter notion has been true for all-miRNA transdifferentiation studies (see Direct Lineage Reprogramming section).

Table 1. Known miRNAs that modulate induced pluripotency
miRNAMembersFunction of miRNATargets testedFunction of miRNA targetReference
  1. a

    ND, not determined.

  2. b

    miR-17 family consists of three miRNA clusters: miR-17–92 cluster (miR-17, miR-18a, miR-19a, miR-20a, miR-19b-1, and miR-92a-1); miR-106a–363 cluster (miR-106a, miR-18b, miR-20b, miR-19b-2, miR-92a-2, and miR-363); and miR-106b–25 cluster (miR-106b, miR-93,and miR-25).

  3. Abbreviations: ESC, embryonic stem cells; ESCC, ESC-cycle-regulating miRNAs; iPSC, induced pluripotent stem cell; miRNA, micro RNA; MEF, mouse embryonic fibroblast; MET, mesenchymal epithelial transition; OSKM, Oct4, Sox2, Klf4, and c-Myc; TFs, transcription factors; TDH, l-threonine dehydrogenase.

iPSC-inhibitory miRNAs
Let-7 familylet-7a, −7b, −7c, −7d, −7e, −7f, −7g, and −7iInduced upon differentiation; resists induction of pluripotency; opposes ESCC miRNAsc-Myc, n-Myc, Sall4, Lin28, Prdm1, Ezh2Associated with tumorigenesis and/or induction and maintenance of pluripotency; Lin28 and c-Myc repress let-7. Ezh2 and Prdm1 are developmental genes opposing pluripotency[ [13, 20]
miR-34 familymiR-34a, −34b, and −34cA major p53 downstream effector; highly induced upon OSKM delivery; antagonizes pluripotencyNanog, n-Myc, Sox2, SIRT1Associated with pluripotency and/or tumorigenesis; SIRT1 supports pluripotency and repress differentiation[ [18]
miR-9Lowly expressed in pluripotent cells; highly induced upon neuronal differentiation of stem cells; decreases reprogramming efficiencyTDHTDH is involved in pluripotent cell metabolism[ [59]
miR-21Expressed in starting MEFs and highly induced following OSKM introduction; resists induction of pluripotencyp85α, Sprty1p85α represses p53, promoting induced pluripotency; Sprty1 inhibits Erk signaling, facilitating induction of pluripotency[ [19]
miR-29aExpressed in starting MEFs and highly induced following OSKM introduction; resists induction of pluripotencyCDC42, p85α, Sprty1CDC42 and p85α repress p53, promoting induced pluripotency; Sprty1 inhibits Erk signaling, facilitating induction of pluripotency[ [19]
miR-199a-3pA p53-regulated miRNA; highly expressed in starting cells; induces G1 arrest via p21NDb[ [30]
iPSC-promoting miRNAs
miR-302–367 clustermiR-302b, −302c, −302a, 302d, and −367A pluripotency-associated intronic miRNA cluster; belongs to ESCC miRNAs; regulates unique cell cycle of pluripotent cells, MET, and chromatin statusMBDII, MeCP1-p66, MeCP2, AOF1, AOF2 (LSD1), RHOC, TGFβRII, NR2F2MBDII, MeCPs, and AOFs block the opening of the chromatin; TGFβRII and RHOC resist MET; NR2F2 inhibits OCT4, impairing pluripotency[ [21, 36, 56]
miR-370–373 clustermiR-371, −372, and −373A human pluripotency-associated miRNA cluster; belongs to ESCC miRNAs; regulates unique pluripotent cell cycle, MET, and chromatin epigenetic status; miR-372 facilitates reprogrammingMBDII, MeCP1-p66, MeCP2, AOF1, AOF2, RHOC, TGFβRIIMBDII, MeCPs, and AOFs block the opening of the chromatin; TGFβRII and RHOC resist MET[ [21]
miR-17 familyamiR-17–92 cluster; miR-106a–363 cluster; and miR-106b–25 clusterA pluripotency- and germline-associated miRNA family; plays critical roles from preimplantation to postimplantation embryogenesis; contains ESCC miRNAs; miR-17–92 cluster comprise the most important of this family in development and diseaseTGFβRII, Wwp2, Fbxw7, p21, p53TGFβRII resists MET; p53-p21 axis resists reprogramming and ensures the generation of higher-quality iPSCs; the ubiquitin ligases Wwp2 and Fbxw7 target Oct4, Klf5, and c-Myc[ [16, 39]
miR-290∼295 clustermiR-290, −291, −292, −293, −294, and −295A murine pluripotency- and germline-associated miRNA cluster; contains ESCC miRNAs; opposes let-7 familyTGFβRII, Lats2, Akt1, Cdkn1a, Pten, Zfp148, Hivep2, Ddhd1, Dpysl2, Cfl2, 9530068E07RikTGFβRII resists MET; Lats2 antagonizes pluripotency; Akt1 impairs murine pluripotency; Cdkn1a blocks cell cycle progression; Pten blocks PI3K signaling, thereby antagonizing reprogramming; Zfp148, Hivep2, Ddhd1, Dpysl2, Cfl2, and 9530068E07Rik provide barriers to the early stage of reprogramming[ [13, 15, 22]
miR-130/301/721 familymiR-130, −301, and −721A broadly (ESC- and differentiated cell-) expressed miRNA family; modulates hypoxia pathway via targeting DDX6Meox2A developmental gene that antagonizes iPSC induction[ [49]
miR-200 familymiR-200c/141; miR-200b/200a/429An epithelial- and pluripotency-associated miRNA family; suppresses tumorigenesis via inducing cell-cycle arrest and/or apoptosis; promotes METZEB1, ZEB2, Snai1, p53ZEBs and Snai1 antagonize MET; p53 resists induction of pluripotency[ [33, 34, 57]
miR-29bA tumor- and pluripotency-associated miRNA induced by Sox2 at the initial stage of reprogrammingDNMT3a, DNMT3bInhibited early (by miR-29b) but induced late in reprogramming; provides an early barrier to iPSC generation; are required for multi-lineage differentiation potential of pluripotent cells[ [14]
miR-519aA primate-specific miRNA belonging to ESCC miRNAs; synergizes with c-MYC to enhance reprogramming efficiency via stimulating proliferationTGFβRIITGFβRII resists MET[ [50]
miR-138Broadly expressed (pluripotent and differentiated cells); modulates stem cell maintenance and differentiationp53p53 resists efficient induction of pluripotency and ensures the genomic integrity of iPSCs; Ezh2 is a developmental gene that opposes pluripotency[ [29]
miR-181 familymiR-181a, −181b, −181c, and −181dLowly expressed in starting MEFs and in pluripotent cells; highly induced in, and promotes, the early phase of OSK-mediated reprogrammingCpsf6, Nr2c2, Marcks, Bptf, Igf2bp2, Nol8, Bclaf1Provide barriers to the early stage of reprogramming[ [15]
miR-182Highly induced during iPSC generationND[ [13]
miR-369A pluripotency-associated miRNAND[ [57]
miR-205 An epithelial-associated miRNAND[ [33]
miRNAs in lineage reprogramming
miR-1, −9, −133, and −499myo-miRs highly induced upon cardiac differentiation of ESCs; play important roles in cardiac muscle; along with (for human cells) or without (for mouse cells) TFs, convert somatic cells into cardiomyocyte-like cellsND[ [60]
miR-9 and −124 Lowly expressed in pluripotent cells; highly induced upon neuronal differentiation of stem cells; play crucial roles in neurogenesis; convert somatic cells into neuron-like cellsND[ [61, 62]
image

Figure 4. miRNAs mediate critical functions of OSKM over the course of reprogramming. (A): Co-operation of OSKM with miRNAs in induced pluripotency. OSKM represses reprogramming barriers either directly (transcriptionally) or indirectly (post-transcriptionally via miRNAs), thereby gradually transitioning the somatic cell gene regulatory network into an ESC-like transcriptional program. (B): miRNAs modulate critical events in iPSC generation. By targeting critical players enriched in starting somatic cells, pluripotency-associated miRNAs regulate all important events such as mesenchymal-epithelial transition, metabolic reprogramming, and ESC-like cell cycle program during the pluripotent reprogramming of cells. Activations are indicated by arrows and inhibitions by blunted lines. The thickened blunted line indicates stronger inhibition. Abbreviations: Dnmts, DNA methyltransferases; ESCC, ESC-cycle-regulating miRNAs; ERK, Extracellular signal-regulated kinases; iPSC, induced pluripotent stem cells; miRNA, micro RNA; OSKM, Oct4, Sox2, Klf4, and c-Myc; TDH, l-threonine dehydrogenase.

Download figure to PowerPoint

Metabolic Reprogramming, Culture Conditions, and miRNAs

During iPSC generation, a metabolic reprogramming has been identified that transitions parental cells' oxidative metabolism into an ESC-like glycolytic state in the resultant iPSCs [63, 64]. Interestingly, miRNAs modulate the metabolic reprogramming of iPSCs. miR-9 is highly expressed in MEFs compared to iPSCs. It directly targets l-threonine dehydrogenase (TDH) [59], a mitochondrial enzyme that hydrolyzes l-threonine into acetyl-CoA and glycine which fuel pluripotency. Loss of miR-9 function improved, while its gain of function blunted reprogramming efficiency via TDH modulation [59]. Although TDH appears to be the main miR-9 target in induced pluripotency [59], other validated miR-9 targets such as SIRT1 [65] may have fine-tuning functions therein.

Hypoxia, which stimulates glycolysis, has been reported to induce a human ESC-like transcriptional signature in several cancer cell lines, upregulating NANOG, OCT4, SOX2, KLF4, c-MYC, and the miR-302 cluster [66]. It was also shown to activate the miR-302 cluster and other pluripotency TFs during iPSC production [67]. These findings explain, at least partially, why iPSC generation is enhanced under hypoxic conditions [68]. Interestingly, hypoxia was able to dedifferentiate ESC/iPSC-differentiated cells back into a pluripotent state highly similar to ESCs in several pluripotency features including the miRNA profile [69]. These data suggest the importance of culture conditions for altering cell fate. In support of this notion, the fatty acid butyrate was shown to upregulate pluripotency TFs as well as the miR-302–367 cluster, thereby highly improving iPSC generation [70, 71]. Additionally, ascorbic acid (vitamin C) which has a documented effect on miRNAs [72-74] has been demonstrated to improve reprogramming efficiency [75].

More importantly, ascorbic acid prevents aberrant silencing of the imprinted Dlk1/Dio3 locus via indirect interference with Dnmt3a binding to DNA, thereby promoting the successful generation of entirely iPSC-derived adult mice (“all-iPSC” mice) [74]. Dlk1/Dio3 locus is a gene cluster that encodes tens of noncoding RNAs, including several miRNAs (e.g., miR-134, miR-369-3p, and miR-136), proper activation of which being associated with the degree of iPSC developmental potency [76]. iPSCs with aberrantly silenced (hypermethylated) Dlk1/Dio3 locus show poor chimera contribution and fail to support all-iPSC mouse development [76]. However, in bona fide iPSCs as with ESCs, the locus is normally maternally expressed, making full-term development of all-iPSC mice possible [74, 76]. Thus miRNAs in this locus might function to support a normal organismal development. Proper activation of Dlk1/Dio3 locus during reprogramming is also facilitated by other miRNAs. miR-29b, which is induced in the early phase, facilitates Dlk1/Dio3 activation via repression of Dnmt3a and Dnmt3b [14]. Dnmts display low expression levels in the early phase but are upregulated at the later phase of reprogramming [14, 42, 77] and are crucial for iPSCs' multilineage differentiation potential [77]. Through downregulating Dnmts, miR-29b would induce global DNA hypomethylation, allowing several important genes, including the Dlk1/Dio3 locus, to become properly activated. In addition, miR-138 which directly represses p53 has also been shown to support the OSK(M)-based derivation of iPSCs with normal Dlk1/Dio3 activation [29]. These findings imply that by targeting key reprogramming roadblocks, miRNAs catalyze somatic cell transition into iPSCs with full developmental potential.

Induction of pluripotency essentially benefits from extrinsic signals, for example, bone morphogenetic protein (BMP) and fibroblast growth factor (FGF) signals that are supplied in reprogramming culture media. Interestingly, miRNA maturation is regulated by these key signaling axes, leading to differential processing of numerous miRNAs [78, 79]. Thus during cell state transitions, miRNA regulation is not only controlled by exogenous and endogenous pluripotency TFs but also by extrinsic signaling pathways which ripple through the GRN to properly fine-tune its structure for successful induction of pluripotency. It remains to be determined which and how exactly miRNAs are controlled by these pathways during reprogramming.

Direct Lineage Reprogramming

Direct lineage reprogramming or transdifferentiation is the direct transformation of one differentiated cell type into another without passing through a pluripotent stage [1]. While TFs have been the most sought-after molecular candidates for reprogramming, miRNAs have recently gained a large amount of interest as potent regulators of cell fate reprogramming. With various molecular targets and involvement in most biological pathways, miRNAs appear effective in facilitating or inducing cell fate conversions between differentiated cells.

Reprogramming to Neuronal Cells

Neurons are among the pivotal therapeutically relevant cell types and express numerous miRNAs. The miRNAs miR-9 and miR-124 comprise a significant fraction of the neuronal miRNAome and potentiate the neuronal state (reviewed in [80]). They are highly induced on neuronal differentiation of stem cells, and when misexpressed, cause neuronal differentiation of neural progenitor cells [81]. Their ectopic expression has also been observed to trigger a morphological and/or transcriptomic switch toward neuronal program in other cells [82, 83]. Thus, these miRNAs may have the potential to force somatic cells to become neurons.

miR-124 in conjunction with two key neurogenic TFs MYT1L and BRN2 was recently shown to promote full neuronal reprogramming of human somatic cells into stable neuronal-like cells [62]. Strikingly, this conversion, which was the first report on induced-neuron generation from “human” somatic cells, was achieved in the absence of ASCL1 which has frequently been present in diverse neuronal reprogramming cocktails. Furthermore, another study found that a five-factor cocktail consisting of miR-124+miR-9+ASCL1+MYT1L+NEUROD2 could convert human cells into neuron-like cells capable of firing repetitive action potentials with a reprogramming efficiency of approximately 10%. Notably, the three neurogenic TFs in the cocktail were unable to induce reprogramming in cells without miRNAs, which underscores the potency of miRNAs in switching cell fate [61].

miR-124 has also been confirmed to repress a splicing regulatory factor known as pyrimidine-tract-binding protein (PTB) [84] which is highly expressed in non-neuronal tissues and antagonizes the neuronal fate partly through inhibiting neuronal PTB [85] and miR-124 [84]. Recently, a seminal work revealed that PTB knockdown alone in several cell types is both sufficient and efficient to trigger their (trans)differentiation into neuron-like cells [86]. PTB deficiency activated key neurogenic TFs and miRNAs (e.g., miR-9 and miR-124) and downregulated antineuronal proteins including the RE1-Silencing Transcription factor (REST) complex. In this setting, PTB was shown to not only perform its previously documented role of regulating alternative splicing but also to affect mRNA stability via modulating miRNAs. It antagonized the neuronal miRNAs by direct competition on miRNA-targeting sites within the target mRNAs and facilitated the targeting of certain miRNAs on specific transcripts by regulating local secondary RNA structures. These collective studies highlight the pivotal contribution of miRNAs (along with RNA-binding proteins) to the post-transcriptional regulation of gene expression in cell fate decisions.

Reprogramming to Cardiomyocytes

Muscle-specific miRNAs are known as myo-miRs. Recently, transient cotransfection of four cardiac-enriched myo-miRs (miR-1, miR-133a, miR-208a, and miR-499) or miR-1 alone (with or without a JAK inhibitor) was shown to switch the fate of cultured cardiac fibroblasts toward cardiomyocyte-like cells [60]. More importantly, lentiviral delivery of these miRNAs into a damaged myocardial area reprogrammed approximately 1% of resident cardiac fibroblasts into cardiomyocyte-like cells in situ. However, these miRNAs could only partially reprogram tail-tip fibroblasts into immature cardiomyocyte-like cells. This incomplete reprogramming might have resulted from either the insufficiency of the miRNA cocktail or of the transient delivery approach to faithfully activate the cardiac program in extracardiac cell types. Inclusion of additional cardiogenic miRNAs/factors along with other cardiac-promoting small molecules, such as ascorbic acid [87] and inhibitors of Wnt signaling [88], might promote efficient cardiac transdifferentiation. Of note, although another study failed to reproduce these results in mouse cells, miR-1 and miR-133 could aid TBX5, GATA4, HAND2, and MYOCARDIN to optimally transdifferentiate cultured human neonatal and adult cells into cardiomyocyte-like cells, a small fraction of which displayed spontaneous beating [89]. Overall, these preliminary results provided the proof-of-concept that cardiac lineage reprogramming was possible by strategies that implicate miRNAs along with or without cardiogenic TFs. Future improvements in these miRNA-mediated approaches will help derive transdifferentiated cardiomyocytes for drug screening, disease modeling, and cell-replacement therapies.

Application of miRNAs in Regenerative Medicine: What Lies Ahead?

Although miRNAs enriched in ESCs, mature neurons, and cardiomyocytes have thus far been used to reprogram different somatic cells into iPSCs, neuron-like, and cardiomyocyte-like cells, respectively, there are a large number of other clinically relevant cell types/subtypes which should be considered for production via miRNA-mediated reprogramming along with or without other promoting parameters (Fig. 5). To supply an adequate number of cells for therapeutic transplantation, terminally differentiated cells reprogrammed from somatic cells should be produced in large quantities. However, due to the lack of cell division, mature cells typically do not survive for long periods of time in culture. Thus, higher priority should be given to the generation of somatic stem/progenitor/precursor cells using defined approaches which probably implicate miRNAs. This is not unlikely given that somatic stem cells (also known as tissue stem cells) express and are dependent on unique sets of miRNAs [4, 90, 91]. Importantly, because somatic stem cells, unlike ESCs and iPSCs, do not form teratoma, they provide a safer source for cell therapy. Another important issue is the potential to conduct transdifferentiation in situ which would eliminate the need to transplant reprogrammed cells.

image

Figure 5. Potential of miRNAs in “human” somatic cell reprogramming to clinically useful cell types. Given that specific miRNA sets are expressed in a cell-type or tissue-specific manner, they can be exploited for altering somatic cell fates into the cell types of interest. Green text denotes the work already published in the human context, and black text indicates the suggested strategies for cell fate conversion. Abbreviations: iPSC, induced pluripotent stem cell; miRNA, micro RNA; RPE, Retinal pigmented epithelium; SM, small molecules; TF, transcription factor.

Download figure to PowerPoint

Although we have been able to improve reprogramming efficiency and derive safe (virus- and integration-free) reprogrammed cells, each of these goals has often been accomplished at the expense of each other. Reprogrammed cells produced by viruses or integrative vectors can be routinely used in basic research and disease modeling. Yet, for regenerative purposes, a reliable ultra-safe strategy should be taken, for which miRNAs could be interesting candidates as they (a) are tissue-specifically expressed, (b) control numerous genes and pathways, (c) are small-sized and safely transfectable into the cells, and (d) have been demonstrated as potent drivers of cell-fate reprogramming as discussed previously. Current methods for safe delivery of miRNAs often yield low reprogramming efficacy (particularly in inducing pluripotency) [57] and repeated transient transfection would cause oscillations in gene regulation, compromising efficient reprogramming. To overcome these hurdles, mature miRNAs of interest may be stabilized by certain modifications [92], can be delivered as pre-miRNAs via nonintegrative vectors (e.g., mini-circles or episomal vectors), or can be accompanied by other promoters of safe reprogramming such as effective growth factors, episomally encoded master TFs, improved culture conditions and effective small molecules. Targeting p53-regulated proteins and miRNAs (but not p53 itself) would facilitate reprogramming to pluripotency.

Modulation of proper miRNAs has been found to catalyze reprogramming kinetics [18, 22, 56]. Rapid reprogramming by miRNAs minimizes culture time which is of particular importance because increased culture time can lead to culture adaptation-induced epigenetic artifacts [93-95]. Owing to their cell-type specificity and efficient/rapid catalysis of reprogramming, miRNAs may provide both stringent pluripotency markers and better reporters for selecting higher quality iPSCs (and transdifferentiated cells) [96, 97]. This is very important, as the quality of the cells can substantially affect their utility.

Another important issue is that the reprogrammed cells often retain a memory of their parental origin which may impair their functionality [98]. One alternative to minimize the memory is to induce “naïve” pluripotency in human cells by using miRNAs. Naïve-iPSC generation would overcome the limitations of primed-iPSC usage by providing a pluripotent state with the least differentiation propensity. This is quite possible, considering that key miRNAs are differentially expressed between naïve and primed pluripotent states [99].

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Induced Pluripotency
  5. Conclusions
  6. Author Contributions
  7. Disclosure of Potential Conflicts of Interest
  8. References

miRNAs are differentially expressed and play crucial roles over the course of somatic cell reprogramming. Strikingly, knockout of Dicer or of a single miRNA cluster, the miR-302–367 cluster, completely abolishes the emergence of iPSC colonies. Moreover, specific miRNAs, along with or without other factors, promote iPSC generation and transdifferentiation. These data highlight the potential of miRNAs to be employed as potent drivers of cell fate reprogramming. The field of miRNA-mediated reprogramming is young and several outstanding questions remain unanswered. For example: what is the global importance and expression pattern of miRNAs in lineage reprogramming; which miRNAs can be effective in somatic cell reprogramming to different tissue stem cells; what is the targetome of miRNAs during cell fate transition; and what is the mode of TF-miRNA cooperation during pluripotent- and direct lineage reprogramming? Addressing these questions will be necessary for the maturation of the reprogramming field and along with the progress already made will facilitate and accelerate miRNA utility in the generation of desired cell types for disease research, drug development, and potential cell-based therapies.

Acknowledgments

We thank the members of the Department of Stem Cells and Developmental Biology for discussions. We apologize to the authors whose work has not been covered or directly cited due to space limitations. This study was supported by a grant provided by the Royan Institute.

Author Contributions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Induced Pluripotency
  5. Conclusions
  6. Author Contributions
  7. Disclosure of Potential Conflicts of Interest
  8. References

S.M.: collection and assembly of data, data analysis and interpretation, and manuscript writing; S.A.: data analysis and interpretation and manuscript writing; H.B.: conception and design, administrative support, manuscript writing, and final approval of the manuscript.

Disclosure of Potential Conflicts of Interest

  1. Top of page
  2. Abstract
  3. Introduction
  4. Induced Pluripotency
  5. Conclusions
  6. Author Contributions
  7. Disclosure of Potential Conflicts of Interest
  8. References

The authors indicate no potential conflicts of interest.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Induced Pluripotency
  5. Conclusions
  6. Author Contributions
  7. Disclosure of Potential Conflicts of Interest
  8. References