It is widely recognized that complete reprogramming is accompanied by extensive epigenetic remodeling to prevent differentiation and promote self-renewal [15, 16]. Since somatic cell reprogramming requires global epigenetic changes, little is known about the regulation of rDNA transcriptional activity during this process. Here, we detected rDNA transcriptional activity changing during iPSC and S-iPSC generation.

In this study, iPSCs were used to investigate rDNA epigenetic changes occurring in four-factor-mediated reprogramming. We infected MEFs and S-MEFs with “Yamanaka factors” for iPSC derivation. The established S-iPSC lines expressed pluripotency markers, and the differentiation potential was confirmed by teratomas with all three germ layers (Fig. 1). Next, we compared the different rDNA transcription regulation patterns between iPSC and S-iPSC generation. We found that serum deprivation significantly stimulated rDNA transcription level compared with the control group. By the end of reprogramming, both iPSCs and S-iPSCs had acquired upregulated rDNA transcription-related genes (45S/18S, 18S, UBF, TIF-IA, and RPI), UBF proteins, and highly-demethylated rDNA promoter regions, which were corresponding to R1 ESCs (Fig. 2). Cells were distinguished from parental cells with unmodified rDNA transcription profiles, implying active rDNA transcription.

Undoubtedly, donor cells with various differentiation statuses will influence the reprogramming efficiency [25]. Based on our results, the main differences of rDNA epigenetic changes existed within the primary stage of reprogramming: Day 6 MEFs had the lowest expression level of 45S/18S, UBF, RPI, reduced 18S genes, and UBF protein, and a relatively high rDNA methylation level, indicating a low-level rDNA transcriptional activity during early reprogramming. However, the serum starvation group did not undergo a transient rDNA transcriptional inhibition. Conversely, the rDNA transcriptional level was reactivated directly without fluctuation (Fig. 2). We hold the view that serum starvation promoted rDNA transcriptional activity during the early stage of reprogramming. The phenomenon could be partially explained as follows. First, before retrovirus infection, serum-starved MEFs were recovered from 15 % FBS supply. Fibroblasts were released to enter the cell cycle and started mitosis immediately. The activated cell proliferation resulted in a burst of protein synthesis and rDNA transcription within the initial stage of reprogramming. Second, a previous report confirmed that retroviruses required the disassembly of the nuclear envelope at mitosis in order to enter the nucleus and replicate [26]. Chen et al. [23] found that retrovirus-infected synchronized cells prior to the G2/M peak could facilitate retroviral infection efficiency, thereby improving cell proliferation and reprogramming. Above all, this would help explain our results that cell cycle synchronization stimulated rDNA transcription reactivation throughout the reprogramming process.

iPSCs have a striking resemblance to ESCs, including epigenetic marks and the pattern of gene expression [15, 16]. Reprogramming is likely a stochastic process, and epigenetic resetting is essential to overcome the barrier to pluripotency [27]. Cells preserve an intermediate stage with similar morphology to iPSCs and ESCs but without the expression of core pluripotency markers described as partially reprogrammed cells [2830]. Recent studies have identified these pre-iPSCs by certain classification, such as AP+/Oct4-GFP cells and Thy1+/SSEA1+ cells [29, 31]. Stable cell sorting and culture of pre-iPSCs contributes to exploring the regulation mechanism that unlocks partially reprogrammed cells into a fully reprogrammed state. Acquisition of a pre-pluripotent state is supposed to occur during the early stage of reprogramming. In our study, we used MEFs and MEFs subjected to serum deprivation as donor cells. A significant increase of AP+ colonies was observed at Day 6 in the serum starvation group, compared with Day 6 MEFs (Fig. 1e). During the G2 and S phases, rDNA transcription was actively transcribed, and transcription is maximal during the S and G2 phases [32]. The activated cell proliferation resulted in a burst of protein synthesis and rDNA transcription within the initial stage of reprogramming. Therefore, there was a distinct difference in methylation levels of MEFs and S-MEFs. In the normal induction group, decreased expression levels of 45S/18S, 18S, UBF, RPI, and UBF proteins showed a transient rDNA transcriptional inhibition. On the contrary, expression levels of 45S/18S, 18S, UBF, TIF-IA, RPI, and UBF proteins were remarkably upregulated in the serum pretreated group (Fig. 2b). Most S-MEFs were released to the S phase simultaneously, resulting in a stimulation of rRNA and protein synthesis within the initial stage of reprogramming. After retrovirus infection, rRNA synthesis remained in a persistent active state to support the cells’ unusually accelerated proliferation induced by Yamanaka factor transfection. The rDNA methylation level of Day 6 S-MEFs was 10.71 %, which was dramatically lower than Day 6 MEFs at 21.62 % (Fig. 4). Taken together, Day 6 S-MEFs had relatively high rDNA transcription activation. Increased rDNA transcriptional activity by serum starvation pretreatment may help partially reprogrammed cells overcome the epigenetic barrier, leading to the increase of AP+ colonies. The nucleosome remodeling and deacetylation complex (NuRD) is a transcriptional modulator that integrates ATP-dependent chromatin remodeling and histone modifying activities [33]. NuRD has been shown to be required for regulation of Pol I transcription [12]. Recent research considered that the complete erasure of epigenetic mark Mbd3/NuRD was required to modulate ESC transcriptional heterogeneity and maintain ESC lineage commitment [34]. Knockdown of Mbd3/NuRD was sufficient to maintain the pluripotency of ESCs in the absence of LIF and generate fully reprogrammed iPSCs (AP+/Oct4-GFP+) rather than partially reprogrammed (AP+/Oct4-GFP) cells. However, ESCs lacking Mbd3/NuRD would have a restricted differentiation potential [35, 36]. On the contrary, for overexpression of Mbd3/NuRD, cells were trapped in a partially reprogrammed state due to the established heterochromatic features and the silence of ESC-specific marker genes, including Oct4 and Nanog [31]. Methyl-DNA binding domain protein 2 (MBD2) is a member of epigenetic inhibiting factors and could bind to methylated NANOG promoter regions to suppress transcription of NANOG [37]. Overexpressed miR-302 cluster or decreased MBD2 expression was thought to increase NANOG gene expression in cells progressing toward complete reprogramming [38]. Important epigenetic modifications, including diverse histone modifications, are involved in transcription repression and activation [39]. Histone H3 lysine 9 (H3K9) methylation was discovered as an epigenetic determinant for pre-iPSCs to establish and maintain the epigenetic barrier [40]. Above all, proper epigenetic modulation of partially reprogrammed cells might overcome the transition barrier to full reprogramming.

The transcriptional activity of rRNA genes varies between cell types, metabolism conditions, and specific environmental challenges, indicating that epigenetic features change during development and differentiation. Cells under injured metabolism, such as nutrient starvation, oxidative stress, and cell senescence, have an impaired rDNA transcriptional activity, whereas a positive influence that stimulates growth and proliferation upregulates Pol I transcription [13, 41]. In our study, rDNA methylation level of MEFs, S-MEFs, Day 6 MEFs, and Day 6 S-iPSCs were 26.09, 13.74, 21.62, and 10.71 %, respectively (Fig. 4). Firstly, our data showed that normal MEFs had the highest rDNA methylation level (26.09 %). S-MEFs, released from the G0/G1 checkpoint, were involved in active RNA and ribosome synthesis, leading to a lower rDNA methylation level. Secondly, the methylation downtrend between pre-iPSCs and MEFs (26.09 % vs 21.62 % and 13.74 % vs 10.71 %) showed the continuous upregulation of rRNA gene transcription. Lastly, the rDNA methylation level of iPSCs-A1, S-iPSCs-B3, and R1 ESCs were comparable, confirming that iPSCs and S-iPSCs were in accordance with ESCs for high proliferation and active protein synthesis. Somatic cell reprogramming can be accomplished by a variety of ways, such as nuclear transplantation (nuclear transfer) [42, 43], cell fusion [44], and direct reprogramming to pluripotency [16]. Zheng et al. found that MEFs had the highest rDNA methylation level at 22.57 %, ESCs had the lowest (6.76 %), and cumulus cells were in the middle (13.59 %). After nuclear transfer, a nuclear reprogramming strategy, MEFNT embryos preserved the highest methylation level (15.52 %), compared with CCNT embryos (9.67 %) and ESNT embryos (6.36 %), indicating that those methylated rRNA genes in donor cells were not activated fully [19]. Although our methylation data for ESCs and MEFs were not exactly the same as Zheng et al.’s results, we concurred that pluripotent stem cells had more cordial rDNA transcription activity than MEFs, and cell reprogramming would recover rDNA epigenetic statuses of donor cells to varying degrees. However, there were opposite opinions over rDNA epigenetic remodeling involved in adult cell reprogramming. Xenopus egg extract-mediated nuclear reprogramming has been shown to induce remodeling of chromatin and reprogram gene expression in somatic cells [45, 46]. A previous study showed that egg extract elicited remodeling of the nuclear envelope, chromatin, and nucleolus, and resulted in a rapid and stable decrease of ribosomal gene transcription. The downregulation of rDNA transcriptional activity here was distinct from a stress response [47]. Ling et al. [48] believed that cell programming in fact negatively influenced rRNA synthesis and methylation at rDNA promoters was increased in iPSCs as well as in mESCs compared with MEFs. Taken together, it is profoundly suggested that the distinct rDNA transcriptional phenomena hidden behind these diverse reprogramming process require further investigation. The complicated rDNA epigenetic regulatory mechanisms may not be simplified and idealized as a simplified model.