Epigenetics and the Control of Epithelial Sodium Channel Expression in Collecting Duct

Abstract

In eukaryotic nuclei, genomic DNA is compacted with histone and non-histone proteins into a dynamic polymer termed chromatin. Reorganization of chromatin structure through histone modifications, the action of chromatin factors, or DNA methylation, can profoundly change gene expression. These epigenetic modifications allow heritable and potentially reversible changes in gene functioning to occur without altering the DNA sequence, thus extending the information potential of the genetic code. This review provides an introduction to epigenetic concepts for renal investigators and an overview of our work detailing an epigenetic pathway for aldosterone-signaling and the control of epithelial Na⁺ channel-α subunit (ENaCα) gene expression in the collecting duct. This new pathway involves a nuclear repressor complex, consisting of histone H3 Lys-79 methyltransferase Dot1a, Af9, a sequence-specific DNA-binding protein that binds the ENaCα promoter, and potentially other nuclear proteins. This complex regulates targeted histone H3 Lys-79 methylation of chromatin associated with the ENaCα promoter, thereby suppressing its transcriptional activity. Aldosterone disrupts the Dot1a-Af9 interaction via Sgk1 phosphorylation of Af9, and inhibits Dot1a and Af9 expression, resulting in histone H3 Lys-79 hypomethylation at specific subregions, and de-repression of the ENaCα promoter. The Dot1a-Af9 pathway may also be involved in the control of genes implicated in renal fibrosis and hypertension.

INTRODUCTION

For decades, our understanding of gene regulation was based on the simplistic paradigm that transcription results from transcription factor binding to a linear DNA sequence, with subsequent recruitment of the RNA polymerase to the gene promoter. The idea that all nucleated cells in an organism contain the same DNA but do not all express the same genes suggested other mechanisms for gene control. An explosion of recent work has led to a much more complex and rapidly evolving model that involves large multi-protein complexes that serve not only to direct recruitment of components of the transcription machinery but also affect chromatin remodeling through histone modification, DNA methylation, RNA-associated gene silencing, chromosomal inactivation, and genomic imprinting. Since these “epigenetic” controls (so termed because they change gene activity without altering the DNA sequence and lead to modifications that can be transmitted to daughter cells) can vary over time and among cells types, epigenetics offers potential explanations for how tissue-specific gene expression is initiated and maintained, and how environmental influences and aging impact the individual’s genetic background to produce variations in health and disease (1). Recent genome-wide studies of epigenetics, so-called “epigenomics,” using microarray, refinements in chromatin immunoprecipitation (ChIP), and high-throughput sequencing technologies have enabled analysis of unique epigenetic signatures and pathways in greater detail, and have enabled the high-resolution profiling of post-translationally modified histones or histone variants and transcription factor occupancy sites throughout the genome. Together with new advances in methods, such as chromatin conformation capture (2) and in vivo chromatin imaging (3), these studies have broadened our understanding of the long distance DNA interactions and the structure, organization, and regulation of chromatin. Epigenomic studies are also being performed on larger human populations, proving additional information about phenotypic plasticity in health and disease, and serving as a rational basis for the design of new therapeutics.

Epigenetic Mechanisms

Modifications of DNA, microRNAs, and chromatin are major epigenetic mechanisms controlling gene expression.

DNA methylation

In vertebrates, DNA methylation involves addition of a methyl group to the 5th carbon of the cytosine ring catalyzed by DNA methyltransferases (DNMTs) in the context of the sequence 5′-CG-3′ (CpG). CpG is notably under-represented in the human genome, occurring at only ~20% of the expected frequency. However, many regions (~1kb long) have a high frequency of CpG, so-called CpG “islands.” Methylation patterns of CpG islands are essential for gene expression: tissue-specific gene expression is partly related to different methylation patterns of CpG islands, and aberrant CpG island methylation patterns are often found in tumors (4). DNA methylation is typically involved in transcriptional silencing (Figure 1). Promoter methylation may be dynamic, as it is during development, and sometimes is targeted specifically to the bodies of active genes. Because DNA methylation is preserved during DNA replication, it can be inherited. Although it remains controversial whether methylation of germ-line DNA in mammalian genomes is stably inherited, de novo methylation events are often observed in a tissue- and disease-specific manner during somatic cell development and differentiation. For example, kidney-specific expression of human organic cation transporter 2 is regulated by methylation of the proximal promoter, interfering with the transactivation by upstream stimulatory factor 1 (5). Tissue-specific expression of the uric acid transporter 1 gene is coordinated by transcriptional activation mediated by HNF1α/HNFβ heterodimers and repression by DNA methylation (6). In contrast, promoter hypermethylation may silence tumor suppressor genes in cancers. In renal cell carcinoma, for example, promoter hypermethylation of several genes has been implicated in disease pathogenesis (7).

Figure 1. Epigenetic controls on gene transcription.

Transcriptionally silent genes commonly have dense DNA methylation (indicated by starbursts), a closed chromatin structure, interaction with heterochromatin proteins, and histone modifications such as methylation (Me) or phosphorylation (P) of histone tails, which are also typically deacetylated. Transcriptionally active genes usually have sparse DNA methylation, an open chromatin structure, interaction with euchromatin proteins, such as ATP-dependent unwinding enzymes (ATP), and histone modifications such as acetylation (Ac), that allow access of the DNA to the general transcription factors and various co-activator proteins.

MicroRNAs (miRNAs) are a novel and growing class of endogenous, small, noncoding RNAs that negatively regulate gene expression involved in cellular proliferation, apoptosis, differentiation, and other important processes through degradation or translational inhibition of their target mRNAs. In the aggregate, miRNAs may directly regulate up to a third of the genes in the human genome. Nuclear transcription of miRNA genes leads to generation of double-stranded RNAs, which are important for generation of the tightly packed, transcriptionally inactive heterochromatin and gene silencing. The RNase III-like enzyme, Dicer, cleaves these double-stranded RNAs into 21–23 nucleotide small-interfering RNAs (siRNAs). The siRNAs unwind and then incorporate into the RNA-inducing silencing complex (RISC), and the strand complementary to the target mRNA is incorporated into the RISC. An endonuclease present with the RISC either degrades or inhibits translation of specific mRNA targets, leading to gene silencing [reviewed in (8)].

miRNAs have diverse temporal and quantitative expression profiles (9), and are known to reside in operon-like, local genomic clusters. Differential miRNA expression in renal cortex and medulla has been reported (10), and miRNAs have been implicated in renal disease models and acute transplant rejection (11–14). miR-192, which is highly expressed in kidney, down-regulates E-box repressors important for TGF-β-induced Col1a2 expression in diabetic glomerulopathy models (14). As another example, the transcription factor E2F3 regulates the Oncomir-1 oncogenic cluster of miRNA located on chromosome 13. The E2F3-Oncomir-1 axis is activated in Wilms tumor (12). Combined miRNA expression profiling with gene signature analysis or proteomic techniques has been used to identify miRNA-protein interactions perturbed in disease states.

Chromatin regulation and Dot1a

DNA strands wrap around histone octamers, forming nucleosomes that are organized into chromatin, the building blocks of chromosomes. The nucleosome is the basic repeating unit of chromatin and includes two copies of each of the four core histones H2A, H2B, H3, and H4 wrapped by 146 bp of DNA (Figure 2). Each core histone is composed of a structured domain, the histone fold, and two tails. Histone tails are necessary for establishing transcriptionally repressive chromatin and for interactions between nucleosomes [reviewed in (15)]. DNA packaging into chromatin compacts the DNA into the nucleus, but also presents a barrier for regulatory factors to interact with the DNA. Reversible changes in chromatin organization influence gene expression: genes are actively transcribed when the chromatin is open, but silent when the chromatin is condensed. The net effects of nucleosome assembly, stability, mobilization, and disassembly influence gene transcription. Replication-coupled nucleosome assembly, which places dimeric subunits behind the replication fork or at sites of active processes that mobilize pre-existing nucleosomes, maintains transcriptionally silent chromatin. Nucleosome-positioning sequences, ATP-dependent nucleosome remodellers, histone chaperones, post-translational modifications and histone variants regulate nucleosomal stability (16).

Figure 2. Schematic model of a nucleosome and histone H3.

Nuclear DNA is compacted and condensed in chromatin. The basic building block of chromatin is the nucleosome, which contains histone octamers. The histone tails are subject to a variety of post-translational modifications, including methylation by methyltransferases such as protein arginine methyltransferase 6 (PRMT6), acetylation by acetyltransferases such as histone acetyltransferase (HAT) or Sirt1, and phosphorylation by kinases such as protein kinase C-related kinase 1 (PRK1). Dot1a specifically methylates Lys79 in the globular domain of histone H3. These histone modifications often exert epigenetic control of target gene transcription. For the purpose of illustration, the positions of the modified amino acids are not drawn to scale.

Reversible and site-specific histone modifications can occur at multiple sites through acetylation, methylation, phosphorylation, ubiquitylation, sumoylation, ADP ribosylation, deimination, and proline isomerization (Figure 1). Dynamic regulation of these modifications, position- and modification-specific protein interactions, and biochemical crosstalk between modifications leads to further complexity in choreographing transcription (17). In addition to the histone-marking [also known as the “histone code” (18)] model of epigenetic inheritance, dynamic histone turnover facilitates continuous access to DNA-binding proteins that may also propagate active chromatin. How chromatin-modifying enzymes engage their nucleosomal substrates in vivo remains an important question. Protein “code readers” bind to specific sets of histone modifications leading to downstream effects, such as gene transcription, DNA synthesis and repair. Certain protein domains bind modified histones (19). For example, the bromodomain binds acetyl-lysine, and the chromodomain binds methyl-lysine.

Histone acetyltransferases catalyze addition of acetyl groups to the N-terminal tail of a histone octamer. This modification reduces the affinity of histones for DNA and allows RNA polymerase and transcription factors to access the promoter (Figure 1). In contrast, histone deacetylases remove acetyl groups and generally allow transcriptional repression. There is often an interplay between DNA methylation and histone acetylation/methylation. Transcriptionally active regions are typically unmethylated DNA with acetylated histones, whereas repressed regions typically include methylated DNA with deacetylated histones (Figure 1). Histone methylation occurs on arginine and lysine residues. Histone arginine methylation is typically involved in activation of genes in which methylases are recruited as co-activators. The CARM1/PRMT1 (co-activator-associated arginine methyltransferase, protein arginine methyltransferase) family of histone methyltransferases (HMTs) functions in this manner and targets histones H3 or H4 (20). In contrast, the SET (suppressor of variegation, enhancer of zeste, and trithorax) domain family of lysine HMTs silences genes. The Suvar39 enzyme was the first such enzyme described; it was found to methylate histone H3 Lys-9 and to be recruited by co-repressors (21). More recently, a third class of HMTs represented by yeast Dot1 (disruptor of telomeric silencing) and its human homologue, DOT1L, were described (22, 23). These HMTs lack a SET domain and affect gene silencing by methylation of Lys-79 in the histone H3 globular domain (Figure 2). Dot1 was originally identified as a mediator of telomeric silencing in yeast and is highly conserved from yeast to humans (24). It has since been shown to be important for meiotic checkpoint control during the cell cycle (25) and for localization and association of Sir silencing proteins in yeast (26).

Mono-, di- and trimethylation states of lysines on histones often lead to different functions. Histone H2B monoubiquitination by Rad6/Bre1 affects H3 Lys-79 trimethylation by enhancing synthesis of all H3 Lys-79 methylation states (27). Moreover, charge-based interactions of basic patch residues of the histone H4 N-terminus and an acidic patch at the Dot1 C-terminus are important for histone H3 Lys-79 methylation in a trans-histone regulatory pathway (28). Recruitment of Dot1 or DOT1L to methylate histone H3 Lys-79 has been associated with both gene activation and repression. Similar bimodal effects on transcription have been attributed to histone H3 serine-10 phosphorylation (29). Using direct sequencing of ChIP DNA to map human genome-wide distributions of histone modifications and chromatin protein target sites, Barski et al. (30) reported that monomethylations of H3 Lys-27, H3 Lys-9, H4 Lys-20, and H3 Lys-79 correlate with gene activation, whereas trimethylations of H3 Lys-27, H3 Lys-9, and H3 Lys-79 are linked to repression. Indeed, we found that overexpression of Dot1a-EGFP constructs in mIMCD3 cells increased methylation of histone H3 Lys-79 but not H3 Lys-4 (31) and basally repressed ENaCa transcription (32). In contrast, another group found that DOT1L and MLL, an H3 Lys-4 methyltransferase, preferentially co-occupy the proximal transcribed region of active genes, correlating with enrichment of di- and trimethylated H3 Lys-79 and trimethylated H3 Lys-4. Thus, coordinate recruitment of these two methyltransferases might demarcate enhancer regions (33). Interestingly, the MLL1 gene can undergo translocations with partner genes, including Af9, that encode proteins that bind to DOT1L and Dot1a, but lack the native H3 Lys-4 methyltransferase domain. Thus the Dot1a-Af9 complex we describe for the control of ENaCa transcription favors H3 Lys-79 methylation at the expense of H3 Lys-4 methylation, and thus represses ENaCa transcription. Combinatorial interactions of various histone methyltransferases and demethylases likely dictate the “ON” or “OFF” mode of transcription for specific genes.

In early work, our laboratory cloned the mouse Dot1 gene, characterized its genomic organization, and in vivo expression (31). The gene contains 28 exons on chromosome 10qC1, with exons 24 and 28 further divided into two and four sections respectively. Alternative splicing in exons 3, 4, 12, 24, 27 and 28 gives rise to five Dot1 mRNAs (Dot1a-Dot1e). Dot1a is closest to its human counterpart DOT1L, sharing 84% amino acid identity across its 1540 amino acids. The 1114 amino acid protein mDot1b is truncated at its N- and C-termini and contains an internal deletion. Dot1c-Dot1e are incomplete at the 5′-end. The functional importance of Dot1b-e is unknown. Northern analysis with probes corresponding to the methyltransferase domain or the Dot1a-coding region detected 7.6 and 9.5 kb transcripts in multiple mouse tissues, but only the 7.6 kb transcript was evident in mIMCD3-collecting duct cells.

Dot1a and the Epigenetic Control of ENaCα Gene Expression

ENaC plays a major role in Na⁺ reabsorption in the distal tubule, and hence the regulation of Na⁺ balance, extracellular fluid volume, and blood pressure (34), as evidenced by the findings of ENaC mutations associated with the genetic hypertensive and hypotensive diseases, Liddle’s syndrome (35) and pseudohypoaldosteronism type 1 (36). ENaC is expressed in the apical membrane of salt-absorbing epithelia of kidney, distal colon, and lung airways, where it constitutes the rate-limiting step in active Na⁺ and fluid absorption. ENaC consists of three homologous subunits—α,β, and γ—encoded by the Scnn1a, Scnn1b and Scnn1c genes.

ENaC is an important molecular target of aldosterone. Increased circulating mineralocorticoids increase extracellular fluid volume and blood pressure, in large part via activation of ENaCs (37). Aldosterone has complex, and temporally distinct actions on ENaC expression, with both immediate (< 3h) effects, attributed to increased trafficking or activity of ENaC in the apical membrane, and delayed actions (>3 h) that involve the synthesis of new ENaC subunits (38–40). In the cortical collecting duct, aldosterone administration or hyperaldosteronism induced by a low-Na⁺ diet increases ENaCα gene transcription, without increasing β- or γ-subunit expression (41), and this response appears to be rate-limiting for ENaC activity in this segment. Moreover, aldosterone does not alter ENaCα mRNA turnover (37). Hence, the focus of our work has been the control of ENaCα transcription.

Under basal conditions, ENaCα gene transcription is constrained, but it can be induced not only by aldosterone, but also other stimuli, including the immediate early gene serum- and glucocorticoid-induced kinase-1 (Sgk1) even in the absence of steroids (42). Sgk1 responds quickly to changes in serum aldosterone levels, which can rapidly change over a narrow range in response to minor, reciprocal changes in dietary salt intake. Sgk1 can rapidly promote increased apical density of ENaC channels by disinhibiting Nedd4-2-triggered internalization and degradation of ENaCα subunits.

The biological actions of aldosterone on target gene transcription have generally been attributed to aldosterone’s interaction with the mineralocorticoid receptor (MR) and the liganded receptor’s binding to hormone response elements (HRE) in the ENaCα promoter. However, the available data suggest that both HRE-dependent and -independent mechanisms control ENaCα transcription, and tissue-specific transcriptional controls on the gene are operative (43). Aldosterone also induces Sgk1 transcription through MR/HRE-dependent mechanisms (44).

The mechanisms underlying this basal repression and its release during gene induction have not been established. We hypothesized that chromatin-based transcriptional repression and de-repression mechanisms might operate coordinately with the classical pathway of aldosterone-mediated transactivation of the ENaCα gene by liganded nuclear hormone receptors. We discovered such an epigenetic pathway that involves combinatorial interactions of Dot1a, the DNA-binding protein Af9 (45), and Sgk1. Because mIMCD3 cells retain phenotypic properties of the inner medullary collecting duct in vivo, respond to aldosterone, and express the components involved in the novel aldosterone-signaling network reviewed here, we have used mIMCD3 cells as our model. We manipulated Dot1a expression by overexpression or by RNA interference to analyze signaling events involved in basal and aldosterone-stimulated ENaCα transcription. We hypothesized that Dot1a-mediated histone methylation at the ENaCα promoter alters local chromatin structure so as to suppress ENaCα transcription, and that suppression of Dot1a activity allows ENaCα transcription to proceed (Figure 3). Consistent with our hypothesis, overexpression of EGFP-tagged Dot1a resulted in hypermethylation of histone H3 Lys-79 at the endogenous ENaCα promoter, repression of endogenous ENaCα mRNA expression, and decreased activity of a stably integrated ENaCα promoter-luciferase construct. The histone H3 Lys-79 hypermethylation and repression of ENaCα promoter activity required Dot1a’s methyltransferase activity, since methyltransferase-dead mutants did not cause these effects (46). The Dot1a effect was substantial, because knockdown of Dot1a resulted in at least a tripling of endogenous ENaCα expression and of the activity of a stably integrated ENaCα promoter-reporter gene (46). We next postulated that one action of aldosterone to enhance ENaCα transcription is to relieve the Dot1-mediated basal repression. Indeed, aldosterone treatment resulted in downregulation of mDot1a mRNA levels in mIMCD3 cells, histone H3 Lys-79 hypomethylation in bulk histones and at specific sites in the ENaCα 5′-flanking region, trans-activation of the stably integrated ENaCα promoter-luciferase construct, and increased expression of endogenous ENaCα mRNA (46).

Figure 3. Model of Dot1a-Af9-mediated transcriptional repression and aldosterone- and Sgk1-induced de-repression of the ENaCα gene.

Under basal conditions, the histone H3 methyltransferase Dot1a and the DNA-binding protein Af9 are complexed with chromatin associated with specific regions of the ENaCα 5′-flanking region. This presumably facilitates the ability of Dot1a to hypermethylate Lys79 of histone H3, leading to a chromatin configuration that suppresses ENaCα transcription. Aldosterone downregulates the expression of both Dot1a and Af9 leading to decreased abundance of the repressor complex. Aldosterone also induces Sgk1, which phosphorylates Ser435 of Af9, causing disruption of the protein-protein interactions of Dot1a and Af9. This results in hypomethylation of histone H3 Lys79 and release of transcriptional repression of the ENaCα gene, contributing in large measure to the effects of aldosterone to increase ENaCα gene transcription. The contributions of the liganded mineralocorticoid receptor and the putative histone H3 Lys79 demethylase (designated by “DM ?” and the dashed arrow) contributing to this dynamic process remain under investigation.

While Dot1a has been shown to have non-sequence-specific DNA binding activity, it had not been reported to have sequence-specific DNA-binding activity that would allow it to interact with the ENaCα promoter. Accordingly, we sought to identify Dot1a protein binding partners that would confer sequence-specific DNA-binding activity for the complex. Using the yeast two-hybrid assay with Dot1a as bait to screen a mouse kidney cDNA library ligated to the GAL4 activation domain, we identified a specific interaction between Dot1a and the protein encoded by Af9, a putative DNA-binding protein (32). This interaction was confirmed by multiple in vitro and in vivo methods, including co-immunoprecipitation assays, mammalian two-hybrid assays, and GST pull-down assays. The two proteins co-localized in the nucleus of mIMCD3 cells. Chromatin immunoprecipitation (ChIP) assays confirmed that Af9 and Dot1a co-associate with chromatin at specific regions of the ENaCα promoter (32). In preliminary reports, we have identified 3 partially conserved Af9 binding sites in the ENaCα 5′-flanking region by DNase I footprinting (Zhang, W., Xia, X., Reisenauer, M., et al., Abstract, see Acknowledgments). Domain mapping of Dot1a and Af9 revealed that Dot1a aa 479–659 and 829–972 are capable, but less than the full-length of Dot1a, to mediate the interaction with Af9 (32). The Dot1a methyltransferase domain (aa 1–416), the putative leucine zipper domain (aa 576–597), and the very C-terminal part (aa 1112–1540) apparently are not necessary for the interaction.

We next tested the functional consequences of the Dot1a-Af9 interaction on basal ENaCα transcription. Our hypothesis that Af9 binds to the ENaCα promoter, and through its interaction with Dot1a, represses ENaCα transcription by increasing the local level of chromatin-associated histone H3 Lys-79 methylation at the ENaCα promoter was confirmed through multiple complementary methods (Figure 3). Overexpression of Af9 resulted in hypermethylation of H3 Lys-79 at specific subregions of the endogenous ENaCα promoter, and repression of endogenous ENaCα mRNA expression and ENaCα promoter-luciferase constructs. Importantly, overexpression of Af9 together with Dot1a synergistically inhibited expression of endogenous ENaCα in mIMCD3 cells. Conversely RNA interference-mediated knockdown of Af9 caused the opposite effects (32). ChIP assays revealed that overexpressed FLAG-Af9, endogenous Af9, and Dot1a are each associated with the ENaCα promoter. Aldosterone inhibited Af9 expression at both mRNA and protein levels in a coordinate manner with Dot1a. Collectively, these data indicated that Dot1a and Af9 form a nuclear complex that, under physiological conditions, associates with specific sites of the ENaCα promoter, leading to targeted hypermethylation of histone H3 Lys-79 and repression of ENaCα transcription (Figure 3). This mechanism appears to be more broadly applicable to other aldosterone-inducible genes, because Af9 overexpression alone or in combination with Dot1a inhibited mRNA levels of three other aldosterone-inducible genes --- connecting tissue growth factor, preproendothelin, and period homolog --- in mIMCD3 cells (32).

We then demonstrated that both Sgk1 and Af9 associate with the ENaCα promoter and that Sgk1-dependent phosphorylation of Ser435 within Af9 reduces the Dot1a-Af9 interaction, inhibiting Dot1a association with the ENaCα promoter, and leading to targeted histone H3 Lys79 hypomethylation at the ENaCα promoter and derepression of ENaCα transcription (47) (Figure 3). The in vivo correlate of these observations was demonstrated in wild type and Sgk1 null mice subjected to manipulation of dietary salt intake. As expected, wild type mice maintained on a low-salt diet exhibited coordinately enhanced levels of renal ENaCα protein expression and Af9 phosphorylation. Conversely, Sgk1 null mice exhibited reduced levels of Af9 phosphorylation and ENaC! mRNA levels when placed on a low-salt diet compared with wild-type mice fed the same diet (47).

Definition of the full complement of proteins and their interaction in the Dot1a-Af9 nuclear complex will be critical to understanding the function and regulation of its activity and of ENaCα transcription. Our most recent work has provided preliminary evidence for functional and physical interaction of the NAD+-dependent deacetylase Sirt1, a homologue of yeast Sir2, with Dot1a in regulating ENaCα gene expression (Zhang, W., Xia, X., Reisenauer, M., et al., Abstract, see Acknowledgments). Sirt1 targets histone and select non-histone proteins, and plays a role in gene silencing, aging and oxidative stress responses (48–50). In mice, we found that Sirt1 is highly expressed along the collecting duct, and that chronic aldosterone treatment inhibits its expression. In mIMCD3 cells, Sirt1 overexpression or treatment with the Sirt1 activator resveratrol dramatically inhibited basal and aldosterone-induced ENaCα promoter activity. Conversely, Sirt1 knockdown or inhibition of its activity with nicotinamide significantly enhanced ENaCα promoter activity. Coexpression of Sirt1 and Dot1a synergistically inhibited ENaCα promoter activity. Overexpression of Sirt1 inhibited endogenous ENaCα mRNA levels, whereas siRNA knockdown of Sirt1 increased them. In ChIP assays, endogenous Sirt1 associated with four regions of the ENaCα promoter known to associate with Dot1a. Aldosterone inhibited Sirt1 association at two of these regions. Sirt1 and Dot1a co-immunoprecipitated from mIMCD3 cells, and when coexpressed in HEK 293-T cells, colocalized in the nucleus. Thus, Sirt1 appears to be a component of the Dot1a repressor complex. Confirmation of these findings and determining whether Sirt1 acts through changes in acetylation or ADP ribosylation of histone and/or non-histone targets are currently being pursued.

Finally, although a demethylase specific for histone H3 Lys-79 has not yet been identified, we speculate that the effect of aldosterone to de-repress ENaCα transcription may involve not only relief of the basal inhibitory effects of Dot1a on the gene, but also recruitment and action, perhaps involving the liganded MR, of a corresponding histone H3 Lys-79 demethylase (Figure 3). It has been proposed that distinct combinations of HMTs and histone demethylases write a histone code for promoter-specific, signal-dependent transcription. In this paradigm, HMTs may impose gene-specific, inhibitory “gatekeeper” functions to prevent unliganded nuclear receptors, such as the estrogen receptor α, and other classes of regulated transcription factors, from constitutively binding to and activating their target gene promoters when activating signals are absent. This “gatekeeper” function is overcome on binding of liganded receptors and recruitment of specific histone demethylases that permit ligand- and signal-dependent gene activation (51).

Potential relevance to health and disease

The relevance of this novel pathway in health and disease is presently open to conjecture, but we agree with Pearce and Kleyman (52) that the Dot1a-Af9 complex as a transcriptional effector mechanism of Sgk1 signaling offers significant efficiency, agility, and capacity in meeting the demands on renal salt excretion in response to acute and chronic changes in dietary salt intake. We also speculate that dysregulation of the Dot1a-Af9 complex may lead to disorders of renal Na⁺ excretion, and, thus, potentially abnormal blood pressure phenotypes in humans, much the same as activating or inactivating mutations in ENaC leads to monogenic forms of Na⁺ retention and hypertension (Liddle’s syndrome) or renal Na⁺ wasting and hypotension (pseudohypoaldosteronism type 1) (34). Moreover, aldosterone has clinical effects beyond its regulation of Na⁺ balance, in particular, the promotion of cardiac and renal fibrosis (53, 54). Since one of the genes we found to be regulated by the Dot1a-Af9 complex, CTGF, promotes cardiac and renal fibrosis in experimental models, we speculate that disturbances that inactivate the Dot1a-Af9 pathway and consequently de-repress CTGF transcription could accelerate heart and kidney pathology. Given the growing interest in epigenomics and the increased technological sophistication, it is likely that epigenetic pathways will be discovered that govern the regulation of a large array of genes of relevance to renal physiology and disease.