Epigenetics and cell cycle regulation in cystogenesis

Abstract

The role of genetic mutations in the development of polycystic kidney disease (PKD), such as alterations in PKD1 and PKD2 genes in autosomal dominant PKD (ADPKD), is well understood. However, the significance of epigenetic mechanisms in the progression of PKD remains unclear and is increasingly being investigated. The term of epigenetics describes a range of mechanisms in genome function that do not solely result from the DNA sequence itself. Epigenetic information can be inherited during mammalian cell division to sustain phenotype specifically and physiologically responsive gene expression in the progeny cells. A multitude of functional studies of epigenetic modifiers and systematic genome-wide mapping of epigenetic marks reveal the importance of epigenomic mechanisms, including DNA methylation, histone/chromatin modifications and non-coding RNAs, in PKD pathologies. Deregulated proliferation is a characteristic feature of cystic renal epithelial cells. Moreover, defects in many of the molecules that regulate the cell cycle have been implicated in cyst formation and progression. Recent evidence suggests that alterations of DNA methylation and histone modifications on specific genes and the whole genome involved in cell cycle regulation and contribute to the pathogenesis of PKD. This review summarizes the recent advances of epigenetic mechanisms in PKD, which helps us to define the term of “PKD epigenetics” and group PKD epigenetic changes in three categories. In particularly, this review focuses on the interplay of epigenetic mechanisms with cell cycle regulation during normal cell cycle progression and cystic cell proliferation, and discusses the potential to detect and quantify DNA methylation from body fluids as diagnostic/prognostic biomarkers. Collectively, this review provides concepts and examples of epigenetics in cell cycle regulation to reveal a broad view of different aspects of epigenetics in biology and PKD, which may facilitate to identify possible novel therapeutic intervention points and to explore epigenetic biomarkers in PKD.

Introduction

Irrespective of the mutations of genes associated with polycystic kidney disease (PKD), the progression of PKD is highly variable between individuals [1]. With a family history, a large percentage of autosomal dominant PKD (ADPKD) patients are expected to develop cysts over their lifespan [1]. However, the course and rate of progression of renal cysts are unpredictable with prognostic tools currently available in the clinic, and the degree of cyst progression is variable [2]. The molecular basis for this variability remains poorly defined. In addition, the rate of cyst progression is also variable in ADPKD patients even within the same family with the same genetic mutations of a specific PKD gene, suggesting that there are non-genetic factors that influence the progression of cystic disease. The study of how non-genetic factors, including patient age, sex, body composition, diet, exercise and microbiome, act upon the genome to influence gene expression and phenotype intersects with the field of epigenetics.

Epigenetics is the study of heritable genome wide changes in phenotype or gene expression caused by mechanisms other than changes in the underlying DNA sequence [3], which enable us to interrogate the mechanisms that underlie disease phenotype and shed new light on the basis for interpatient variability in disease progression. Similar to the genetic information found within the sequence of DNA, epigenetic information can be inherited across generations, transmitted from mother to daughter cells, and is required for life [4]. The epigenetic mechanisms include DNA methylation, histone modifications, and noncoding RNA-associated silencing [5]. Epigenetic activation and silencing are ways to turn genes on and off, which may explain, in part, why genetic twins are not phenotypically identical [5]. DNA methylation at specific loci and in whole genome can now be quantified in a sequence-specific manner across the entire genome to generate a methylome map, and can be quantified in either single cells or circulating free DNA [6–8]. The advanced approaches to the study of histone modifications should highlight the functional associations of alterations in the chromatin landscape with disease progression.

The roles of epigenetic modulation on gene expression and protein function in PKD have become the focus of scientific investigation [9, 10]. Recent studies suggested that inherited PKD gene mutations in patients may favor the development of epigenetic changes that increase the progression of renal cysts [9, 10]. An interactive picture between PKD gene mutations and the epigenome needs to be further developed. Epigenetic regulators may be one of the modifiers that regulate the initiation and progression of PKD. Thus, a deeper understanding of the epigenetic changes associated with PKD should lead to an improved understanding of mechanisms involved in cyst initiation and development, which can ultimately be applied towards development of strategies for clinical management.

In this review, we define the term of “PKD epigenetics” and group the PKD epigenetic changes into three categories, including PKD Epigenetic modifiers, PKD epigenetic mediators and PKD epigenetic modulators, which helps PKD investigators to easily understand the roles of PKD epigenetics in each category. We focus on the epigenetic mechanisms in PKD and the interplay between epigenetic mechanisms and cell cycle regulation which leads to cystic renal epithelial cell proliferation and cystogenesis. In addition, we discuss the prospective role of DNA methylation as a biomarker in PKD.

The term of “PKD epigenetics”

A multitude of genetic studies, ranging from candidate-gene studies to genome-wide association studies, have identified a number of genetic susceptibility factors for PKD and its clinical phenotypes, but the contribution of these factors to renal cyst susceptibility is only modest [11]. Therefore, in addition to providing a comprehensive description of the current understanding of genetic susceptibility underlying PKD, it has been proposed that epigenetic mechanisms are involved in the regulation of the expression and activity of PKD genes and the components of PKD associated signaling pathways during cyst development [9, 10]. The studies of the epigenetic mechanisms in PKD have been defined as “PKD epigenetics”.

PKD epigenetics is the study of somatically heritable changes in DNA methylation, histone modifications and genes encoding epigenetic modifiers by comparison between normal cells and cystic cells with the application of the advanced epigenetic techniques. The expression of PKD genes and PKD associated genes and the function of PKD associated proteins can be regulated by epigenetic modifications, including DNA hypermethylation and hypomethylation [5,6] and histone (H2a/2b/H3/H4) modifications of nucleosomic octomers [12], which can often be distinguished from their unmodified counterparts using both standard molecular biology techniques and next-generation genome-wide sequencing approaches. Methods, such as bisulfite treatment for investigating DNA methylation [13] and chromatin immunoprecipitation (ChIP) and ChIP-sequencing for investigating histone modifications and protein-chromatin interactions [14], are extremely powerful to determine mechanisms of epigenetic regulation and the importance of epigenetics to different aspects of gene expression and cell biology of PKD. However, epigenetic measurement, high-throughput sequencing of DNA and RNA, data acquisition, and analysis require equipment and skills that lie beyond the scope of most investigators’ laboratories interested in PKD. Thus, we provide a detail discussion about the advanced epigenetic techniques that can be applied in PKD epigenetic study in our recently publication [15].

The complexity of epigenetic mechanisms that function either locally at the gene level or globally across the epigenome presents significant challenges in PKD epigenetics. However, the constant technique improvements in next generation sequencing, genome-wide quantitative data acquisition and analysis are helping to meet this challenge [15]. The highly dynamic nature of epigenetic mechanisms in response to PKD gene mutations and non-genetic factors offers hope for epigenetic therapies in cystic diseases, as is now occurring in clinical oncology.

The categories of PKD epigenetic changes

The PKD epigenetic changes can be grouped into three categories: epigenetic modifiers, epigenetic mediators, and epigenetic modulators, as has been grouped in cancer.

PKD Epigenetic modifiers are the proteins that directly modify the epigenome, e.g., DNA methylation, post-translational modification of histone, or higher-order chromatin structure. Most of the proteins altered by mutation in PKD are epigenetic modifiers, supporting that both genetic and epigenetic changes are connected through the epigenome. The epigenetic modifiers that have been associated with PKD pathogenesis include DNA methyltransferases (DNMTs), histone deacetylases (HDACs), histone methyltransferases (HMTs) and bromodomain proteins [16–21] (Figure 1).

Figure 1. Classification of PKD associated factors based on their role in PKD epigenetics.

The already known PKD regulators are grouped into three categories: epigenetic modifiers, epigenetic mediators and epigenetic modulators. Epigenetic modifiers are at the center of PKD epigenetics, including all the DNA and histone modification enzymes, which target downstream epigenetic mediators directly or indirectly and are modulated by the upstream epigenetic modulators to form a positive feedback loop. PKD gene mutations and non-genetic factors stimulation in general result in the up- or downregulation of epigenetic modifiers mediated by epigenetic modulators and then to affect the expression and function of epigenetic mediators. In this figure, only known PKD epigenetic modifiers, epigenetic mediators and epigenetic modulators are included. We expect more factors will be added in the future.

PKD epigenetic mediators are downstream direct targets of the epigenetic modifiers or the downstream targets of epigenetic modification by the modifiers, and their alteration contributes to cystic phenotypes. The epigenetic mediators identified and associated with PKD pathogenesis include p53, retinoblastoma protein (Rb)/E2F, NF-κB, STAT3, HSP90, tubulin, and hedgehog signaling components, etc. [16–19, 21, 22] (Figure 1).

PKD epigenetic modulators are the upstream factors of the epigenetic modifiers, which influence the activity or localization of the epigenetic modifiers and the epigenetic states mediated by epigenetic modifiers. These modulators bridge the genetic mutations of PKD genes with the epigenome, whose disordered function confers a predisposition to and acceleration of cyst development. The epigenetic modulators associated with PKD pathogenesis include Rb/E2F (on the expression of DNMT1 and EZH2 (one of the HMTs), p21/PCNA (proliferating cell nuclear antigen) (on the interaction with DNMT1), AMPK (on the phosphorylation of DNMT1), HEF1/Aurora A (on the phosphorylation and activation of HDAC6), NF-κB/STAT3 (on the expression of Smyd2 (one of the HMTs), etc. [18, 23–26] (Figure 1).

Epigenetic mechanisms in PKD

Abnormal epigenetic regulation may alter gene expression and function, which leads to diseases such as ADPKD. ADPKD, in essence, is a genetic disease [1]. However, recent studies indicate that cyst development cannot be accounted for by genetic alterations alone, but also involve epigenetic changes such as DNA methylation, histone modifications and microRNAs [8, 10, 20, 27]. These three epigenetic mechanisms affect the patterns of gene expression that regulate cystic phenotype, such as cystic renal epithelial cell proliferation and apoptosis, renal inflammation and fibrosis growth [10, 27].

DNA methylation and PKD

DNA methylation is the first recognized and the most common and widely used mechanism for epigenetic modifications in cells, which controls gene expression in the “off” position [28]. DNA methylation is predominantly found on CpG sites of the mammalian genome, called “CpG islands” [29]. During evolution, the CpG dinucleotide has been selectively depleted through conversion of methylated cytosines to thymidines via a deamination process [30]. Only 10% of human genome contains CpG sites, and about 70 to 80% of those CpGs are methylated [30]), primarily in heterochromatic regions. Genomic DNA methylation is regulated by DNA methyltransferases (DNMTs: DNMT1, DNMT3a and DNMT3b), which mediate the transfer of methyl groups from S-adenosylmethionine to the 5′ position of cytosine bases in CpGs [31]. The addition of methyl groups by DNMTs to CpG sites attracts methyl-DNA-binding proteins, resulting in repression of gene activity [32]. There are only small regions of DNA (1 to 2%) are rigorously protected from methylation and are associated with the transcription start sites in almost half of human genes [33].

In general, DNMT1 functions to maintain DNA methylation in mammalian cells and is therefore responsible for accurately replicating genomic DNA methylation patterns during the S phase of the cell cycle [34], whereas DNMT3a and DNMT3b functions for the de novo methylation of DNA [35]. However, it has also been found that both groups of enzymes exhibit some levels of maintenance and de novo methylation in vitro, suggesting that this classification of the DNMTs may be oversimplified [36].

DNA methylation status has high stability and serves as a special epigenetic memory and marks of specific cells throughout all periods in the cell cycle [37]. Inside the cells, S-adenosyl methionine acts as an important methyl group donor, in which folic acid and B12 play the determinant roles in re-methylation or the attraction of de-methylated form of S-adenosyl methionine through passive and active mechanisms [37]. Treatment with folic acid and B12 could change DNA methylation patterns and alter the levels of gene expression [38]. Aberrant DNA methylation patterns, including hypermethylation and hypomethylation compared to normal tissue, have been associated with a large number of human cancer and diseases [39]. All types of cancers have been shown to possess global genomic hypomethylation, yet at the same time, local areas may be hypermethylated. Typically, there is hypermethylation of tumor suppressor genes and hypomethylation of oncogenes [40, 41]. Similar to cancers, ADPKD genome also exhibits global hypomethylation compared with non-ADPKD kidneys [20], whereas the global hypomethylation on ADPKD genome does not exclude the possibility that local areas may be hypermethylated at the same time. To support this possibility, two groups reported that PKD1 is either hypermethylated in whole gene-body regions or methylated at the 3’ end of the gene body in ADPKD genome [20, 42]. We found that PKD1 mutations result in the upregulation of DNMT1 in cystic renal epithelial cells (unpublished data), which further suggests that some genes downstream of PKD mutations (including ADPKD and ARPKD genes) may be hypermethylated in specific cell type(s) during cyst formation. Although the methylation of other ADPKD or ARPKD genes is undetermined, however, they may also be changed in the genome of ADPKD and ARPKD patients. The role of the abnormal DNA methylation on PKD genes in the manifestation and progression of PKD need requires further investigation. In addition, to identify the changes of methylation on the whole genome and on specific loci with advanced next generation sequencing will help us to understand the general roles of DNA methylation in PKD. Furthermore, the identified aberrations of methylation on specific gene(s) may have potential as diagnostic markers in the progression of PKD.

Histone and non-histone modifications in PKD

Histone and non-histone modifications are the second recognized epigenetic mechanism that regulates gene expression and protein function [43, 44]. In mammalian cells, 2 m of DNA is condensed via interaction with histones to form tightly packed chromatin [45]. The basic unit of chromatin is known as the nucleosome, which consists of 147 bp of double stranded DNA wrapped around a complex of eight histone proteins (two copies each of H2A, H2B, H3 and H4) and the linker DNA between the nucleosome core particles [45]. Each of the core histones has an unstructured N-terminal amino acid tail extension that can be acetylated, methylated on lysine and arginine, phosphorylated on serine, ubiquitinated, sumoylated and ADP-ribosylated by epigenetic modifiers to control chromatin structure and function [46]. Histone modifications are highly dynamic processes that add or remove post-translational modifications by different epigenetic modifiers, respectively [43]. For gene transcription, histone post-translational modifications can serve as marks for recruitment of chromatin remodeling complexes and transcription factors, activation of transcriptional enhancers, recruitment of repressive proteins, and interaction with the DNA methylation machinery [47]. In general, histone acetylation loosens chromatin to transcriptionally active euchromatin and constitutes binding scaffolds for bromodomain-containing protein complexes [43], in which a BET Bromodomain protein, Brd4, has been associated with PKD [48]. Histone methylations, on the other hand, are correlated with either gene activity or repression depending on which histone residue is modified. For instance, trimethylation of lysine 9 of histone 3 (H3K9me3) and H3K27me3 are associated with condensed and transcriptionally silent heterochromatin, whereas trimethylation of lysine 4 of histone 3 histone (H3K4me3) and trimethylation of lysine 20 of histone 4 (H4K20me3) are generally associated with euchromatin [49]. In addition to post-translational modifications of nucleosomal histones, some of the epigenetic modifiers also have the capacity to modify non-histone substrates, such as 1) transcriptional factors of p53 [50], Rb [51], STAT3 [18], and 2) tubulin [52, 53], protein-folding chaperone HSP90 [54], and estrogen receptor α [55], etc.

PKD gene mutations result in the up- or down-regulation of epigenetic modifiers that regulate cell proliferation, differentiation, apoptosis, and inflammation [56, 57]. The involvement of epigenetic modifiers in histone/lysine modification in these processes has been identified in the past few years [10, 18]. It has been reported that, 1) HDAC6 and SIRT2, a class III HDAC, regulate the disassembly of cilia [21, 26], and fluid flow regulates the phosphorylation and function of HDAC5 [58], which integrates the “fluid flow-cilia hypothesis in PKD” to epigenetic regulation; 2) HDAC6 regulates epidermal growth factor receptor (EGFR) endocytic trafficking and degradation in renal epithelial cells [59]; 3) HDAC inhibition targets Id2-Rb/E2F signaling to regulate cystic epithelial cell proliferation [19]; 4) Sirtuin 1, another class III HDAC, regulates cyst development through deacetylation of Rb and p53 in ADPKD [16]; 5) a BET bromodomain protein, Brd4, regulates cyst growth in mouse models of ADPKD through c-Myc-p21 signaling [48]; and 6) Smyd2 regulates cyst development through methylation of STAT3 and NF-κB in ADPKD [18]. The epigenetic modification can occur in other PKD genes and epigenetic regulators may also regulate signaling pathways other than those listed above, including HSP90 [60], STAT6 [61–63], AMPK [64], Wnt/β-catenin [65], ILK/mTOR [66, 67], hedgehog [68], NF-κB/inflammation signaling [69, 70], etc. For example, the signaling pathways, including GSK3β and notch signaling, have been associated with epigenetic modifier(s) [71–73]and may regulate renal cyst progression. Thus, determining epigenetic modifications of PKD genes and establishing interactive connections between epigenetic modifiers and PKD associated signaling pathways is highly significant for our understanding of the pathogenesis of PKD, which should be constantly investigated.

Noncoding RNAs (ncRNAs) contribute to epigenetic mechanisms in PKD

Over the past decade it has become apparent that is of critical importance for, as discussed in greater detail elsewhere [74, 75]. Noncoding RNAs generated from the nonprotein coding fraction of the human genome play critical roles for homeostasis and disease [76], which are currently divided into several classes based on their length and the processing and effector mechanisms [77]. MicroRNAs (miRNAs) are the most studied class of noncoding RNAs, which control about 60% of the transcriptional activity of protein encoding genes. MicroRNAs are about 22 base long ribonucleotide sequences that target complementary untranslated regions of mRNAs to direct them for degradation in the RNA-induced silencing complex, or to regulate their translation [78]. Other noncoding RNA types include, small nucleolar RNAs (60 to 300 bp size) which are involved in ribosomal RNA modifications [79], the PIWI interacting RNAs (24 to 30 bp size) which interact with PIWI proteins that are critical for genome stability regulation [80], and large intergenic RNAs and long noncoding RNAs (>200 bp size) which are found in chromatin complexes [81]. Noncoding RNAs are considered part of the epigenetic machinery due to their critical involvement in epigenetic phenomena [82]. For example, long noncoding RNAs can recruit chromatin remodeling complexes to specific loci, and are involved in DNA methylation and other chromatin modifications [83]. Much of the current work in this field has been directed towards understanding of the roles and mechanisms of miRNAs which have also been shown to play key roles in PKD [27]. The roles of miRNAs in PKD and their therapeutic potential are discussed in another review in this special PKD edition.

Interplay between epigenetic mechanisms and cell cycle regulation in PKD

Epigenetic DNA and chromatin modifications, which are dynamic across the cell cycle (Figure 2), play an essential role in regulating cell-cycle progression locally by controlling the expression of individual genes and globally by controlling chromatin condensation and chromosome segregation [84], whereas epigenetic machineries can also be influenced by cell-cycle progression. The mammalian cell cycle includes four successive phases in order: G1 (the postmitotic interphase), S (the DNA synthetic phase), G2 (the postsynthetic phase) and M (mitotic phase) [85] (Figure 2). Any defects, such as DNA damage, DNA replication stress and chromosome segregation abnormality, may induce cell cycle arrest at various stages to allow cells to repair the defects [86–88]. Abnormalities in gene expression and protein function related to cell cycle regulation may result in deregulated cell proliferation as has been characterized in cystic renal epithelial cells [89]. The Rb/E2F and p53/p21 signaling are well studied pathways that control cystic renal epithelial cell proliferation [16, 19], which have been characterized to not only drive or prevent cell cycle progression [90, 91] but also function as epigenetic mediators or as targets of epigenetic modifiers, including DNMT1, HADCs and HMTs, etc. [51, 92–94]. The discussion of the crosstalk of PKD associated epigenetic modifiers, including DNMT1, Sirt1, bromodomain protein, Smyd2 and EZH2, and cystic cell related cell cycle regulator(s), including Rb/E2F, p53, Id2, p21 and PCNA (Figure 3), should not only increase our understanding of the roles of epigenetics in PKD but also may provide an example for PKD investigators to consider the integration of the signaling pathways they are focusing on with epigenetic mechanisms.

Figure 2. Major features of chromatin and epigenetic mark changes during the cell cycle.

Cells in G1 phase exhibit active gene transcription, such as PCNA, with 1) increased global acetylation of histones H2A and 2B, H3, and H4; 2) decrease DNMT1 and DNMT3 and global DNA methylation; AND 3) disruption of HDAC–Rb/E2F–Smyd2(?) complex. During S-phase histones are transcribed and synthesized, DNA is replicated and new and recycled nucleosomes assemble to form nascent chromatin. In S phase, DNMT1 and global DNA methylation as well as methyltransferases for histone H3 and H4 and methylation of histone H3 and H4, are increased, and acetylations of histone H3 and H4 are conserved on existing and nascent chromatin. During G2 phase, nucleosomes are mature and histone biogenesis is inhibited. During M phase, chromosomes condense and many transcription factors and chromatin binding proteins are disassociated from the chromatin. Histone H3 phosphorylation as a mitotic hallmark is increased, and global deacetylation of histone H2A and 2B, H3, and H4 and high methylation of H3K9, H3K17, HAK79, H3R3 and H4K20 are observed. Green arrow heads represent nuclear pore complexes.

Figure 3. The crosstalk of PKD associated cell cycle regulators with epigenetic machinery.

Cell cycle regulators identified in PKD pathogenesis include Rb/E2F, p53/p21, Id2, c-Myc, PCNA, etc., and the epigenetic modifiers identified in PKD pathogenesis include DNMT1, Class I and II HDACs, SIRT1 and SIRT2, histone methyltransferases Smyd2 and EZH2, bromodomain protein, Brd4, etc. A close interaction among these factors has been identified. The Rb/E2F pathway plays a critical role in the connection of most of the PKD associated cell cycle regulators. Rb is a main repressor of E2F-mediated transcriptional activity via recruiting of HDACs, DNMT1 and HMTs to the E2F-site-containing promoters (indicated at the center of this figure). Disruption of the Rb/E2F-HDACs complex, Rb/E2F-DNMT1 complex, or Rb/E2F-HMTs (EZH2, Smyd2, etc.) complex may release this repression, which may increase E2F mediated the expression of DNMT1 and EZH2 (top left corner of this figure). On the other hand, the expression and activity of p21, which can be regulated by p53, Id2 and Brd4-c-Myc signaling (bottom left of this figure), may inhibit the activity of Cdk2/4 to affect the phosphorylation of Rb and also the binding of PCNA with DNMT1 to a affect DNA methylation (bottom right corner of this figure). Smyd2 can methylate not only RB and p53 to affect their activity (middle right of this figure) but also histones to directly regulate gene expression (top right corner of this figure).

The dynamics of epigenetic marks and chromatin modifications during the cell cycle

Progression of cell cycle can be regulated by epigenetic modifiers either locally by activating and repressing cell cycle regulated genes or globally by controlling chromatin condensation and chromosome segregation, which in turn ensures correct inheritance of the epigenetic marks to the new daughter cells. In G1-phase and the G1–S transition, large amounts of genes, including the E2F family of transcription factors known as key players with Rb, termed Rb/E2F, in the mammalian cell cycle [95, 96], are under active transcription accompanied by global acetylation of nucleosomes mediated by histone acetyltransferase. At the same time, the levels of DNMT1 and DNMT3b was decreased [97], resulting in a decrease in global DNA methylation levels [98]. Rb is a main repressor of E2F-mediated transcriptional activity, which forms a complex with HDACs, DNMT1 and SUV39H1, a histone methyltransferase (HMT) and recruits them to E2F-site-containing promoters, such as cyclin A/E and Cdk2, etc. [99, 100]. We and others found that Rb could be methylated by one of the HMTs, Smyd2 [16, 18, 51, 101], and this methylation increased it phosphorylation which might by mediated with Cdk2/4. Whether Smyd2 is recruited by Rb to E2F-site-containing promoters needs to be further investigated. This repression of RB on E2F can be relieved by the disruption of the HDAC–Rb–SUV39H1 complex with the phosphorylation of Rb by cyclin D-CDK4/CDK6 at the end of the G1-phase [95], and we propose that treatment with the inhibitors of HDACs and Smyd2 should decrease cystic renal epithelial cell proliferation by inhibition E2F signaling (Figure 2). At this phase, the active epigenetic marks H3K36me3, H3K79me2 and H3K4me3 occupy the active promoter of PCNA, whereas the repressive mark H4K20me3 does not bind on the promoter of PCNA [102].

In S phase, the transcription of genes is low, whereas the levels of DNMT1 and global DNA methylation are increased. Global methylation of histone H3 and H4 as well as the activity of corresponding histone H3 and H4 methyltransferases are also increased in late G1 and S phases [103], whereas H4K20 and H3K79 methylations are not detected on newly deposited histones until later time points in the cell cycle [104]. In contrast, the acetylations of histone H3 and H4 are constantly conserved [105]. For instance, H3K9Ac and H4K16Ac, which peaks during G1-phase, remain high during S-phase [103, 106]. Notably, during S phase, high levels of activation marks, such as H3K9/H3K14 acetylation and H3K4me3, but low levels of H3K9me3 are associated with early-firing origins of replication, whereas a compact chromatin conformation marked with low levels of H3K9Ac/H3K14Ac and H3K4me3 and increased levels of H3K9me3 are associated with late-firing origins [107] (Figure 2).

In G2 and M phase, protein synthesis is markedly conducted in G2 phase and histone H3 phosphorylation as a mitotic hallmark is gradually increased (Figure 2). The phosphorylation of histone H3 is initiated in defined chromosomal domains during the G2-phase and spreads throughout the genome on prophase in mitosis in concert with a condensed chromatin state [108]. The phosphorylation of other histones has also been shown to be up-regulated during M-phase and linked to chromatin condensation [109–111]. In M phase, global deacetylation of histone H2A and 2B, H3, and H4 on different lysine sites by different HDACs was observed, which is thought to ensure a correct packaging of nucleosomes into metaphase chromosomes [109–113]. Thus, transcription is believed to be turned off during mitosis. Remarkably, Sirt2, an PKD associated HDAC with preference for H4K16, has been shown to be highly expressed and associated with chromatin exclusively during mitosis [112]. In addition, high levels of the methylation of H3K9, H3K17, HAK79, H3R3 and H4K20 correlated with the deacetylation of histones during M phase [109, 114, 115].

In mature mammals, most cells reside in a prolonged non-dividing state called G0-phase. In cell culture, cells usually enter G0-phase under conditions, such as contact inhibition or serum starvation. When cells enter G0 phase, specific genes, such as cyclin genes, are inactivated, which is mainly linked to the E2F-dependent transcriptional machinery [116, 117]. A general decrease in DNA methylation is observed in serum-starved cells as compared with cycling cells [97, 118], whereas the levels of DNMT1 were also low in G0 phase [118].

Rb and p53 are direct targets and p21 is indirect target of epigenetic modifiers

Rb participates in a well-characterized cell cycle regulatory process that is controlled by cyclin-dependent kinases (cdk) and cdk inhibitors [119]. Rb was known to be inactivated via phosphorylation by Cyclin D-Cdk4/6 complex [120]. It was shown that Rb exists in three states: un-phosphorylated, monophosphorylated, and hyper-phosphorylated. Each has a unique cellular function [121].When a cell enters G1, Cyclin D-Cdk4/6 phosphorylates Rb at a single phosphorylation site. Then, active mono-phosphorylated Rb causes repression of E2F-targeted genes specifically. There are 14 mono-phosphorylated Rb isoforms and each isoform has distinct E2F binding preferences, which suggests that mono-phosphorylated Rb has a diversity of functions [122]. Hyper-phosphorylation of monophosphorylated Rb is mediated by Cyclin E-Cdk2 complex that allows cell cycle to entry into S phase [90]. At the end of mitosis, hyper-phosphorylated Rb is de-phosphorylated by PP1 directly to its un-phosphorylated state. Un-phosphorylated Rb then drives cell cycle exit and maintains senescence [90]. Inactivation of Rb by epigenetic modifications has been associated with increased cystic renal epithelial cell proliferation [16, 18]. We and others found that Rb can be deacetylated by one of the class III HDACs, Sirt1, and can be methylated by one of the HMTs, Smyd2 (Figure 3) [16, 18, 51, 101]. These modifications increased the phosphorylation of Rb, which may be mediated by Cdk4/6 [120], but decreased its activity, leading to increase cystic renal epithelial cell proliferation. Both Sirt1 and Smyd2 can also deacetylate and methylate p53 on lysines, respectively, to impede p53 from binding to its target gene promoters, such as p21, and knockdown of Smyd2 enhances DNA damage-induced and p53-dependent apoptosis [123, 124]. Smyd2 is also able to methylate histone-3 lysine-36 (H3K36) to repress transcriptional activity via its association with histone deacetylase repressor complex [125].

In addition to p53 mediated p21 transcription, the expression of p21 can also be indirectly regulated by a BET Bromodomain protein, Brd4, which recognizes acetylated-lysine residues in histone tails to regulate the expression of numerous genes associated with cell cycle, cell growth, inflammation, and cancer[126–129]. Inhibition of Brd4 in cystic renal epithelial cells with JQ1, a selective small-molecular inhibitor of BET bromodomain (BRD) protein(s), 1) increased the levels of p21 mRNA and protein, which was transcriptionally repressed by Brd4 mediated c-Myc; and 2) decreased the phosphorylation of Rb, leading to decrease cystic epithelial cell proliferation as shown by inhibition of S-phase entry [48]. Furthermore, HDAC inhibition (HDACi) targets inhibitor of differentiation 2 (Id2), thus increasing Id2 mediated p21 transcription, leading to decrease p21 mediated Cdk2/Cdk4 activity and then Cdk2/Cdk4 mediated Rb phosphorylation to downregulate cystic epithelial cell proliferation [19, 130]. These studies suggest that epigenetic modifiers may be through targeting Rb/E2F, p53 and p21 to regulate cystic epithelial cell proliferation (Figure 3).

Rb associates with complexes to modify chromatin and DNA methylation in cystic renal epithelial cells.

Traditional views of Rb function suggest that inactivation of Rb by its phosphorylation leads to deregulation of E2F target genes, such as cyclin E. In this process, Rb acts as a recruiter that allows the binding of proteins to alter chromatin structure onto the site E2F-regulated promoters [131]. The majority of activities attributed to Rb are believed to result from direct interactions between Rb and either HDACs or transcription factors [132–135]. For instance, Rb recruitment of histone deactylases, including HDAC1, HDAC3, and HDAC3, leads to the formation of nucleosomes and their further packing into chromatin at E2F-regulated promoters to repress the transcription of E2F target genes. However, Rb may also function through interactions with components of the DNA methylation machinery, as Rb can regulate the activity of DNMT1 via the direct interaction of Rb and DNMT1 as shown in glutathione S-transferase pull-down and coimmunoprecipitation experiments (Figure 3) [136, 137]. In addition, Rb overexpression greatly diminished the ability of DNMT1 to bind DNA and thereby induced genomic hypomethylation [136]. Furthermore, the expression of DNMT1 can be regulated by RB/E2F pathways [23, 138]. It was found that the expression of DNMT1 was elevated in Rb knockout murine prostate epithelial cell lines, and conserved E2F consensus binding sites were identified in proximity to the transcription initiation sites of murine and human DNMT1. In the absence of pRb, DNMT1 transcripts exhibited aberrant cell cycle regulation and aberrant methylation of the paternally expressed gene 3 (Peg3), a tumor suppressor gene [23]. Most recently, a study examining cell cycle–specific gene expression identified DNMT1 as part of a G₁-S cycle cluster, which included several other E2F target genes, including cdk2, cyclin E, MCM3/5/6, PCNA, replication factor C, DNA polymerase I, and p107 [23]. Thus, there is considerable evidence to support the interplay between Rb and DNMT1 in cell cycle regulation, which also suggests a role of Rb in the methylation of ADPKD genome.

p21 and PCNA associate with DNA methylation in cystic renal epithelial cells.

Overexpression of DNA methyltransferases has been found in cystic renal epithelial cells (unpublished data). Increased expression of DNMTs should be correlated with increased methylation of CpG islands as found within the promoters of tumor suppressor genes [139]. These changes may contribute to genomic instability and PKD gene mutations. In addition to binding with Rb in cell cycle regulation, it has been reported that binding of DNMT1 to PCNA, which is always upregulated in cystic renal epithelial cells and PKD kidneys, coordinates DNMT1 activity and DNA replication and that this step is negatively regulated by the protein p21 (Figure 3) [25]. The binding of p21 to PCNA competes with the site of PCNA to DNMT1 leading to inhibit DNMT1 activity [25]. How might these dynamics mediate normal and abnormal DNA methylation in PKD? Perhaps, in normal cells, p21 negatively regulates the binding of DNMT1 to PCNA, primarily in early S phase, and protects CpG islands from methylation. In cystic cells, loss of p21 function, as we described above, allows increased DNMT1 more access, via PCNA, to DNA replication foci, possibly facilitating aberrant methylation of CpG islands. This study suggests that loss of p21 may contribute to altered DNA methylation in cystic renal epithelial cells, which needs to be further investigated.

The regulation of a histone modifier, EZH2, by the cell cycle machinery

While epigenetic modifiers mediated chromatin regulates cell cycle events, such as the transition between each phase of cell cycle and chromosome segregation at mitosis, the cell cycle machinery also impacts chromatin by regulating the histone modifiers. One example of the cell cycle dependent regulation of the histone modifiers is the regulation of the mammalian homolog of Enhancer of zeste, E(z) in Drosophila, EZH2, by the cell cycle. EZH2 is the major methyltransferase for H3 lysine 27 and plays a crucial role in differentiation gene silencing through interaction with the Polycomb Repressive Complex 2 (PRC2) [140–142]. EZH2 is a direct target of the core cell cycle transcriptional regulator E2F (Figure 3) [24], and is upregulated in cystic renal epithelial cells. Targeting EZH2 with its inhibitor delayed cyst growth in Pkd1 knockout mouse models (ASN abstract, 2017). During cell cycle, EZH2 can be phosphorylated on Thr350 by Cdk1 and Cdk2, which silences differentiation-associated genes, such as silencing of HOX genes and SOX family members [143, 144]. EZH2 can also be phosphorylated by Cdk1 at Thr487, which disrupts the binding of EZH2 to the other PRC2 components, leading to the de-repression of EZH2 target genes, resulting in premature differentiation of human mesenchymal stem cells [145]. The cell cycle associated positive and negative outcomes of EZH2 are important for normal cell proliferation. However, how these outcomes of EZH2 are balanced in actively proliferating cystic cells remains unclear. Thus, in order to fully understand if Rb/E2F regulates EZH2 expression during cell cycle and how elevated EZH2 coordinates with the cell cycle machinery to promote cystic cell proliferation, further investigations will be required.

In sum, extensive connections between epigenetic chromatin modification and the cell cycle machinery clearly exist, which impact gene expression and cystic cell proliferation in multiple ways. In this section, we discuss how epigenetic modifiers, such as HDACs, HMTs and bromodomain protein, regulate the function of cell cycle regulators in cystic renal epithelial cells and then how cell cycle regulators, including Rb/E2F, p53/p21 and PCNA, function as epigenetic modulators to regulate the expression and activity of DNMT1 to affect DNA methylation, and to regulate the expression and activity of epigenetic modifiers during cell cycle. The direct interaction of DNMT1 with EZH2 on the methylation of cell cycle associated genes has been reported [146]. Thus, the interplay among these factors should bring together the fields of DNA methylation, cell cycle regulation, control of chromatin organization, and PKD genetics. Future work will continue to uncover new molecular connections between the cell cycle machinery and epigenetic modifiers, to help us finally understand how epigenetic modifications regulate cystic cell cycle and proliferation.

Prospective role of DNA methylation as a biomarker in PKD

The variable disease course of ADPKD and the late occurrence of functional decline make it important and urgent to develop new prognostic biomarkers and surrogate endpoints. Recently, epigenetic changes including alteration of DNA methylation patterns have been associated in kidney disease progression [20, 42, 147–150]. Following identification of cell-free nucleic acids in systematic circulation, increasing evidence has demonstrated the potential of cell-free epigenetic biomarkers in the blood or other body fluids for cancer detection [7, 151]. Cell-free circulating DNA (cfcDNA) carries distinctive DNA methylation markers in certain GC-rich fragments, which are usually located within the promoters and first exons of many genes, comprising CpG islands [152]. Blood cfcDNA has DNA methylation patterns specific for different diseases [153]. Differential DNA methylation analysis of cfcDNA by bisulfite modification is possible for different types of cancer and benign proliferative and/or inflammatory diseases [152], such as PKD. DNA-based biomarkers developed using cell-free circulating DNA in blood have been used successfully for prenatal diagnosis [154], with applications for cancer detection and diagnosis, and monitoring of treatment efficacy starting to emerge [155–158]. We have performed the whole genome bisulfite sequencing (WGBS) [159] analysis to determine the changes in global DNA methylation in patients with ADPKD compared to healthy subjects, and to identify the target genes with hypermethylated CpG islands in PKD genome (unpublished data), which should provide a specific and sensitive target for PKD treatment. However, if cell-free circulating DNA from ADPKD patients carries distinctive DNA methylation markers in certain GC-rich fragments remains elusive. We propose that subclinical PKD may release DNA molecules into the circulation that bear highly specific epigenetic signatures defining the stage of PKD (Figure 4), as has been proposed in cancer. Thus, further investigation of DNA methylation using cfcDNA should facilitate the identification of accurate biomarkers for diagnosis, prognosis and prediction of response to therapy and ascertainment of outcomes.

Figure 4. Potential applications of PKD clinical samples in epigenetic mechanism study.

Dysregulated DNA methylation can be characterized in biopsy kidneys and in the blood or other body fluids with the advanced epigenetic technologies. The abnormal DNA methylation signatures identified in ADPKD genome should guide us to identify blood based DNA methylation signatures in Cell-free DNA, which is generated from necrotic/apoptotic cells and can be detected in the circulation, potentially providing epigenetic biomarkers that could be used to aid diagnosis, prognosis and prediction of response to therapy and ascertainment of outcomes.

The major challenge to the discovery of cfcDNA methylation biomarkers is the dearth of efficient techniques for methylation analysis. Such techniques have to combine high sensitivity and accuracy with tolerance to inherent heterogeneity of clinical samples and be able to measure methylation at multiple sites within the same sample. However, the rapid evolution of epigenetic technologies that can be directly applied to DNA biomarker discovery and validation meet these challenges.

Combating PKD with Epigenetic Therapy

More and more studies have demonstrated that epigenetic changes are essential factors in cyst initiation and development. Epigenetic abnormalities occurring in PKD should lead to the development of epigenetic treatment in PKD, which aims to reverse the epigenetic alterations occurring in PKD, thus, restoring the normal epigenome. Great progress has been accomplished in epigenetic drugs targeting on chromatin and histone-modifying enzymes. The main targets of recent epigenetic treatment are to inhibit abnormal DNMTs and HDACs using specific inhibitors, in which two DNA methyltransferase enzyme (DNMT) inhibitors, 5-azacytidine and 5-aza-2′-deoxycytidine [5], and a deacetylase (HDACs) inhibitor SAHA have been approved by the FDA as effective drugs for cancer treatment [160]. Meanwhile, other epigenetic drugs, such as FK228, Tacedinaline, Entinostat, nicotinamide and MS-275, have already been tested in phase III clinical experiments [160]. More specific and effective inhibitors should be developed to reduce unwanted side-effects as much as possible since epigenetic modifying enzymes function in a wide range of organs in the body. The epigenetic changes in PKD are constantly identified and abnormal expression of epigenetic modifiers in controlling histone acetylation and methylation has been reported in ADPKD animal models [16]. Targeting histone deacetylases with trichostatin (TSA) or nicotinamide (vitamin B3) and targeting histone methyltransferases, Smyd2 and EZH2, with their inhibitors delayed cyst growth in Pkd1 knockout mouse models [10, 16, 18, 19], which suggested that epigenetic regulators should be included in the clinical management of ADPKD. However, there is still a long way to go for epigenetic treatment of PKD. Research on detailed epigenetic mechanisms in PKD is expected to enhance the ability to diagnose and treat PKD.

Perspectives and conclusions

Increasing evidence suggests that epigenetic changes in DNA methylation, histone modifications, transcription factor binding, and noncoding RNA expression contribute to the pathogenesis of PKD [9, 16, 17, 19–21, 59]. We define the term of “PKD epigenetics” as the study of somatically heritable changes in DNA methylation, histone modifications and genes encoding epigenetic modifiers by comparison between normal cells and cystic cells with the application of the advanced epigenetic techniques. For the foreseeable future, PKD epigenetics will contribute in at least three ways to PKD research. First, the systematic mapping of functional PKD epigenome should generate a rich set of hypotheses to be further tested in order to identify relevant pathways, and to understand phenotypic variation and plasticity in PKD. Secondly, the advances in epigenetic technologies should help to identify epigenetic biomarkers in the blood or other body fluids [7, 151]. We believe that technological developments, technical improvements, and a better understanding of cfcDNA biology are likely to yield targets and assays that improve on previous outcomes of cfcDNA methylation for diagnosis and prognosis as well as for monitoring therapy. Thirdly, the epigenetic drug discovery in human diseases, although in its infancy, may result in identification of novel therapeutic targets in PKD.

The epigenetic mechanisms in PKD and epigenetic technologies that can be applied to PKD research have been discussed in detail in our recent publications [9, 10, 15]. Thus, in addition to include most recent epigenetic discoveries in PKD, we focus on the crosstalk between epigenetic mechanisms and cell cycle regulation in normal and cystic renal epithelial cells in this article. Multiple pathways including p53/p21 and Rb/E2F, and many key epigenetic mediators of cell cycle regulation contribute to PKD pathogenesis. However, how these critical cell cycle regulators are altered is not clearly understood. Recent advances in the study of epigenetics have shown that these pathways may be regulated by epigenetic modifications. Most importantly, these pathways also play critical roles to affect epigenetic mechanisms, including DNA methylation and histone modification. A comprehensive discussion about the interplay between PKD associated signaling pathways on cell cycle regulation and epigenetic mechanisms should provide examples for PKD investigators to consider a connection of a specific signaling pathway that they are focusing on with epigenetic regulation, and vice versa.

Highlights.

In the review article, we define the term of “PKD epigenetics” and group the PKD epigenetic changes into three categories, including PKD Epigenetic modifiers, PKD epigenetic mediators and PKD epigenetic modulators, which helps PKD investigators to easily understand the roles of PKD epigenetics in each category. We focus on the epigenetic mechanisms in PKD and the interplay between epigenetic mechanisms and cell cycle regulation which leads to cystic renal epithelial cell proliferation and cystogenesis. In addition, we discuss the prospective role of DNA methylation as a biomarker in PKD.