3. DNA Methylation
Epigenetic DNA modification (methylation) occurs when a methyl group derived from S
-methionine is bound to position 5 of the cytosine ring, forming 5-methyl-cytosine (5mC) [17
] (Figure 2
). This happens at specific dinucleotide sites called CpGs, as these are prone to such modifications [10
]. CpG islands are short sequences of DNA with a high linear frequency of 5’-CpG-3’ sequences. The “P” denotes the phosphoesteric bond linking the cytosine and the guanine nucleotides [9
]. CpG islands are often located in the promoter regions and approximately 40% of genes contain CpG islands in the end 5’ region (promoter, untranslated region, and exon1). The rest of the genome (intergenic and intronic regions) is considered CpG poor [17
]. In healthy somatic cells, up to 90% of CpG dinucleotide sites are methylated, except for those located in the promoter region, which appear to be somewhat protected from such modification [17
]. Non-CpG island DNA methylation has also been reported to influence protein-DNA interactions, gene expression and chromatin structure, and stability [18
] (Table 1
). Functionally, DNA methylation suppresses gene transcription, so DNA hypermethylation results in gene silencing [10
Different degrees of DNA methylation have been correlated with variable onset, timing, and severity of EH (Table 1
]. Global genomic DNA methylation can be quantified by measuring the amount of 5-methyl cytosine present in a DNA sample. Smolarek et al.
] found a correlation between decreased levels of 5mC in peripheral blood with an increase in EH grade severity. The concentration of 5mC was calculated as a ratio of spot intensities and a significant deficit in 5mC was observed in healthy controls, relative to levels found in the two severity grades of EH. The mean level of 5mC was 1.80 ± 0.69 in healthy controls and 1.14 ± 0.48 in all hypertensive subjects (specifically, 1.29 ± 0.50 grade 1, 0.99 ± 0.42 grade 2), indicating that global DNA methylation levels decrease as the severity of EH increases. However in this study, the percentage of 5mC was expressed as a coefficient “R
” value, and the method for calculating “R
” was not disclosed.
The most robust data on the involvement of methylation in blood pressure regulation comes from the study by Kato et al.
]. Kato and colleagues performed a genome-wide association and replication study and identified genetic variants at 12 new loci that correlated with blood pressure modulation in 320,251 individuals of East Asian, European, and South Asian ancestry [20
]. An investigation of the relationship between the sentinel blood pressure single nucleotide polymorphisms (SNPs) with local DNA methylation in 1904 South Asians (peripheral blood; Illumina Human Methylation 450 Bead Chip array (Erasmus Medical Centre, Rotterdam, The Netherlands)) revealed a two-fold enrichment between sequence variation and DNA methylation. Twenty-eight of the 35 sentinel SNPs were associated with one or more methylation markers [20
], suggesting that DNA methylation may lie on the regulatory pathway linking sequence variation and blood pressure. Genes associated with the 12 newly identified SNP loci include those involved with vascular smooth muscle (IGFBP3
) and renal function (ARHGAP24
]. Investigation of cross-tissue patterns of DNA methylation at the leading 26 CpG sites associated with the sentinel SNPs showed that DNA methylation in the blood was closely correlated with methylation patterns of various tissues (liver, muscle, subcutaneous and visceral fat). While these potentially ground-breaking results indicate that methylation levels in blood may provide a proxy for methylation patterns in other tissues, further research is needed to investigate this correlation with a hypertensive affector organ, such as the kidney.
Initial DNA methylation research centred on correlating EH with the global level of 5mC [19
], but more recent studies have focused on the methylation of specific DNA sequences. The hydroxysteroid dehydrogenase-11β2 enzyme (HSD11B2) converts cortisol to cortisone, which is found in circulatory concentrations up to 1000 times higher than aldosterone (the primary mineralocorticoid hormone and sodium (Na) transport modulator of the renin-angiotensin-aldosterone system (RAAS)) [10
]. Cortisol and aldosterone bind mineralocorticoid receptors with equal affinity, but due to the difference in concentration of the hormones, cortisol has a larger role in sodium reabsorption by the kidneys and, in consequence, plays a larger role in fluid-related arterial pressure [10
]. HSD11B2-mediated degradation of cortisol to cortisone is disrupted when the promoter region of the HSD11B2
gene is hypermethylated [27
]. The resulting imbalance in the active metabolites of cortisol and cortisone, tetrahydrocortisol (THF), and tetrahydocortisone (THE), respectively, promotes the onset of HT [27
], in accord with the findings by Friso et al.
]. Friso et al.
] also found that HSD11B2
promoter methylation was associated with EH, with parallel disruptions to the THF/THE ratio.
Due to the well-known involvement of the RAAS system on arterial pressure regulation, any changes in the activation status of RAAS-regulated genes have a pronounced effect on an individual’s potential to develop HT. This effect has been extensively tested in animal models of HT [29
]. Hypomethylation of the promoter regions of the angiotensin II type 1b receptor gene in the adrenal glands of the maternal low protein (MLP) rat promoted HT through an exaggerated response to salt [29
]. The same effect was seen in a second study, where maternal protein deficiency during pregnancy induced hypomethylation of the promoter regions of RAAS-responsive genes, such as angiotensin converting enzyme (ACE), causing predisposition to HT in offspring as well as cognitive deficits [22
]. The human PRCP
gene encodes a lysosomal prolylcarbopeptidase protein, which is implicated in cleavage of c-terminal amino acids linked to proline in peptides such as angiotensin II and III [30
]. Methylation profiling conducted on young African males revealed that the PRCP
gene was hypomethylated in EH subjects [23
]. Similarly, when expression patterns of the angiotensin 1a receptor (Atgr1a
) were analysed in both spontaneously hypertensive rat (SHR) and its normotensive control, the Wistar Kyoto rat (WKY), expression of Atgr1a
was significantly increased by week 20 in the SHR. Bisulfite sequencing revealed that the Atgr1a
promoter from endothelial cells in the aorta and mesenteric arteries of the SHR rats became progressively hypomethylated with age compared to their WKY counterparts [24
], suggesting heightened Atgr1a
expression in the SHR was related to the hypomethylation of the Atgr1a
promoter and may have a role in the maintenance of high blood pressure. Somatic ACE (sACE) converts angiotensin I to the active form, angiotensin II and is, therefore, a key regulator of blood pressure [25
]. Bisulfite sequencing conducted on cultured endothelial cells and Wistar Kyoto rats revealed that hypermethylation was associated with transcriptional repression of sACE, indicating possible epigenetic involvement of sACE modulation in HT [25
Membrane transporters such as Na+
cotransporter 1 (NKCC1), participate directly with fluid and electrolyte loss and therefore arterial pressure regulation [26
]. NKCC1, expressed by the Sic2a2
gene, mediates the transport of sodium, potassium and chloride into and out of the cells [10
]. Changes in ion fluxes have been implicated in EH, particularly an augmented, passive influx Na+
, and Cl−
in hypertensive vascular smooth muscle cells [26
]. Hypomethylation of the Sic2a2
gene promoter in the SHR aorta and heart, result in increased expression of NKCC1 and is positively correlated with HT [26
Methyl CpG binding protein 2 methylates the norepinephrine transporter gene, silencing its expression [10
]. Hypermethylation of the norepinephrine transporter gene, which transports norepinephrine and dopamine from the synaptic gap back to the pre-synaptic neuron, has been shown to lead to increased transport responsiveness, resulting in EH [31
]. Furthermore, phenyl-ethanolamine N
-methyltransferase (PNMT), which can act similarly to methyl CpG binding protein 2, has been shown to exacerbate the decrease in norepinephrine uptake, thereby enhancing the local and systemic catecholaminergic effect [31
4. Histone Modification and Hypertension
DNA is packaged into the dynamic protein structure of chromatin, whose basic unit is the nucleosome. A nucleosome comprises of two copies each of the histone proteins H2A, H2B, H3, and H4 [32
]. Post-translational modifications regulate gene expression by controlling the dynamics of chromatin. Modifications occurring at residues in histone tails include acetylation and methylation (Figure 3
) as well as phosphorylation, sumoylation, and biotinylation [32
Epigenetic histone modification occurs when the N-terminal tail is subjected to a variety of post-translational modifications [10
]. Up to 60 possible modifications can occur. Specifically, the nucleotide lysine can be modified by methylation, acetylation, ubiquitylation, or sumoylation, whereas arginine can only be modified by methylation, and serine and threonine are modified only by phosphorylation [10
]. While each histone modification pattern has a unique impact on the corresponding nucleosome, generally, the outcomes are similar: histone acetylation promotes gene transcription while deacetylation represses transcription; histone lysine methylation in position 79 inhibits while histone arginine methylation activates gene transcription; histone lysine hypermethylation or mono-methylation at position 9 can have opposing effects—respectively silencing or activating the target gene [9
] (Table 2
Histone modification affecting arterial pressure levels has been documented in a variety of human and animal tissues, including vascular smooth muscle. Vascular oxidative stress can contribute to endothelial dysfunction—a hallmark of HT—and the development of HT. A study by Bhatt et al.
] documented the beneficial effects of resveratrol on endothelial function in the SHR. A later study of the mechanisms behind this effect was conducted by Han et al.
]. Epigenetic modifications, in the form of up-regulated H3K27me3 expression was observed in the renal artery of a salt-sensitive hypertension model of the Wistar rat [33
], possibly correlating with the improvement observed in HT status but this was not clear. Endothelial nitric oxide synthase (eNOS) is primarily responsible for the production of nitric oxide in the vascular endothelium, and plays a key role in the regulation of vascular tone [33
]. eNOS activity is thought to be down-regulated in CVD. Fish et al.
] found that eNOS expression appears to be controlled by cell-specific histone modifications. They also demonstrated that the nucleosomes of endothelial cells corresponding to the eNOS core promoter were enriched in lysine 9 of histone H3 and lysine 12 acetylation of histone H4, and of di- and tri-methylation of lysine 4 of histone H3. Fish et al.
] observed that all of these epigenetic histone modifications were significantly important for the expression of eNOS [34
The study by Riviere et al.
] noting the modulation of sACE by CpG methylation also documented the involvement of histone deacetylases in the same process. Later findings of Lee et al.
] further validated these observations. Riviere et al.
] also found that tissues from SHRs exhibit higher expression levels of Ace1 mRNA and protein than those of WKY controls. They found the Ace1
promoter regions of the SHR tissues were more enriched with H3Ac and H3K4me3, and concluded that Ace1
is locally up-regulated in SHR tissues via histone code modifications [35
In line with observations by Lee et al.
], Cho et al.
] found that levels of the membrane transporter Nkcc1
mRNA and protein in the aortas of Sprague–Dawley (SD) rats were significantly increased after administration of an angiotensin II infusion. Cho et al.
] found that acetylated histone H3 (an activating histone), was significantly increased together with greatly decreased trimethylated histone H3 (a deactivating histone) [36
], suggesting that both histone modification and/or DNA demethylation have a role in the epigenetic up-regulation of Nkcc1
during HT development.
Disruptor of telomeric silencing-1 (DOT 1), a methyl-transferase, enhances methylation of lysine 79 residue of histone H3 (H3K79) [10
]. DOT1-mediated hypermethylation of H3 disrupts silencing of genes associated with maintaining telomere length during DNA repair, located in the telomeric regions of chromosomes [40
]. This disruption correlates with a decrease in connective tissue growth factor transcription, increased intracellular cAMP and alterations in the adaption of blood vessels to stressors—associated with HT [40
]. In Caucasians, DOT1-like histone H3K79 methyltransferase (DOT1L) is strongly associated with increased blood pressure in response to hydrochlorothiazide [37
Activation of the renal sympathetic nervous system has long been thought to play a crucial role in the development of salt-reactive HT [41
]. Overactivity of renal sympathetic pathways can promote the activation of several mechanisms which lead to increased sodium retention, including stimulation of renin release, reduced renal blood flow, and increased sodium reabsorption in the loop of Henle [42
]. Serine-threonine kinases have a key role in renal tubular sodium reabsorption; specifically, the with-no-lysine (WNK) family of serine-threonine kinases [44
]. In normotensive mice, a low-sodium diet decreases expression of renal WNK4, whereas a high sodium diet increases its expression [45
]. Renal sympathetic activity can also promote sodium reabsorption via activation of the β-2 adrenergic receptor (β2
AR) leading to cyclic AMP (cAMP) production, and increased activity of renal epithelium sodium channels (ENaC) [30
]. WNK4 signalling targets ENaC, and may also mediate the prohypertensive effects of renal sympathetic overactivity. In addition to the classical cAMP pathway, cAMP modulates gene transcription via inhibition histone deacetylase-8 activity leading to increased histone acetylation [47
]. Mu et al.
] found that salt loading triggers β-adrenergic-mediated down-regulation of renal WNK4, increases salt retention, and consequently leads to the development of salt sensitive HT. WNK4 was recently found to be negatively regulated by the glucocorticoid receptor [49
]. Mu et al.
] also noted that the underlying mechanism down-regulating renal WNK4 is cAMP dependent modulation of histone acetylation, which consequently increased histone H3 and H4 acetylation. Acetylation of H3 and H4 seems to favour binding of the glucocorticoid receptor to a WNK4
promoter region that contains a negative glucocorticoid-responsive element [38
]. Mu et al.
found WNK4 expression was inhibited via these pathways leading to overexpression of the sodium chloride co-transporter (NCC) and the onset of HT [38
5. Non-Coding RNAs and Hypertension
Non-coding RNAs (ncRNA) are implicated in several epigenetic processes, notably small RNAs that can influence histone modifications and cytosine methylation which are connected with gene expression regulation [50
]. NcRNA modification can produce effects analogous to classical DNA or histone epigenetic mechanisms described earlier, or can induce HT via separate, distinct processes. Essentially, epigenetics allow the same genome to show alternative phenotypes based on different epigenetic influences. Some of the most complicated and researched epigenetic phenomena including X-chromosome inactivation, parental imprinting and paramutation have an RNA contribution, either directly or indirectly [50
]. While most ncRNAs have no identified function yet, it is conceivable that all are participants in epigenetic mechanisms which are yet to be clearly defined. NcRNAs have a strong influence on the “central dogma” of biology. The past view that all DNA was transcribed into RNA and then translated into protein has now been updated to include the fundamental role of ncRNAs on the regulation protein levels. Recently, several small and mid-sized ncRNAs have been described and are known to have a role in the regulation of transcription, post-transcription and translation. Small ncRNAs include PIWI-interacting RNAs (piRNAs), transcription initiation RNAs, and microRNAs (miRNAs) [8
]. Mid-size ncRNAs include small nucleolar RNAs, promoter upstream transcripts, TSS-associated RNAs, and promoter-associated small RNAs. Long ncRNAs (lncRNAs) are most commonly associated with a reduction in transcription but may also have a role in the regulation of miRNA levels [8
]. For a comprehensive review of miRNAs and blood pressure refer to the review paper by Marques et al.
]. Here we will deliver a brief summary of some additional key miRNAs involved in HT as well as a review of lncRNAs.
MiRNAs are the most commonly studied small ncRNA; currently there is no research available investigating the involvement of the other types of small and mid-sized ncRNAs in EH. There are more than 1800 miRNAs in the genome, each approximately 22 nucleotides in length [52
]. miRNAs down-regulate protein-coding genes by binding to target sequences in the 3’ (5’ targeting is uncommon but possible) untranslated region of a target mRNA, resulting in mRNA degradation or repression of mRNA translation [53
] (Figure 4
). Evidence suggests a single miRNA can regulate numerous mRNAs, so it is plausible that miRNAs are causative agents for complex deletions [53
]. miRNAs are increasingly implicated as regulators of cardiovascular system [54
], including modulation of arterial pressure [49
] (Table 3
In regards to the RAAS-regulated genes, it appears that miRNAs may also have a role modulating ACE mRNA transcription [29
]. MiRNA has-miR-155 was found by Sethupathy et al.
] to target the polymorphic sequence located in the 3’UTR of AGTR1
mRNA. This site has a Mendelian C (minor) and an A (major) allele. The minor-C allele is associated with EH [55
]. Has-miR-155 binds most effectively to the A allele, leading to a reduction in AGTR1
mRNA in EH individuals who inherit this allele. Individuals with the C allele did not have reduced AGTR1
mRNA and exhibited a greater pressor effect in response to angiotensin II [55
]. Has-miR-155 is also found to be up-regulated in preeclampsia placentas; it regulates AGTR1
expression in umbilical vein endothelial cells [56
]. A recent genome-wide association study documented expression of mRNAs and miRNAs in kidneys from untreated EH individuals, which were removed as a consequence of renal cancer [57
]. Two miRNAs (has-miR-181a and has-miR-663) with the ability to bind to the 3’UTR of renin mRNA were found to be under-expressed in EH. These miRNAs were able to regulate the expression of a reporter gene and renin-mRNA itself, which explains over-expression of renin mRNA seen in EH kidney [57
]. In reference to DOT1 causing hypermethylation of H3 [37
] (mentioned earlier), small interfering RNAs (siRNA) have been shown to have an opposite effect by silencing the expression of the DOT1
]. Furthermore, siRNAs have been observed to have a role in modulating the RAAS system [61
]. Angiotensin II was shown to induce a contractile response in smooth muscle cells extracted from mice overexpressing p22phox, a subunit of NADPH oxidase responsible for activating NADPH oxidase and generating superoxide anions in atherosclerotic plaques in human blood vessels [62
]. siRNA targeting p22phox inhibits the contractile response from angiotensin II, lowering blood pressure [58
Long non-coding RNAs (lncRNAs), a heterogeneous group of non-coding transcripts longer than 200 nucleotides, regulate their targets by influencing epigenetic control, mRNA processing, translation or chromatin structure. LncRNAs have been implicated in several biological processes, including transcriptional regulation by epigenetic mechanisms [63
]. An exemplary representation is X-chromosome inactivation present in the female genome of mammals which involves the inactivation of one copy of a female sexual chromosome by DNA methylation. This process is coordinated by the X-inactive specific transcript (XIST) and the XIST antisense transcript (TSIX) ncRNAs, very long ncRNAs whose expression can be regulated by epigenetic mechanisms such as DNA methylation [50
Large intergenic non-coding RNAs (lincRNAs) have been identified both as modulators of health development [64
] and disease states [67
], and may regulate the expression of neighbouring genes and distant loci [63
]. LincRNAs are more tissue specific than are protein-coding genes [59
], suggesting they have selective functions in different tissues. Cabili et al.
] have described the annotation and tissue-specific expression of all lincRNAs. Their gene ontology analyses were based on lincRNA expression in different tissues [59
]. The protein-coding genes located close to lincRNA loci included genes previously associated with EH and blood pressure regulation, for example, those encoding corticotropin-releasing hormone [59
], suggesting many lincRNAs have an enhancer-like function that promotes transcription of their neighbouring coding genes [68
]. LncRNAs have a potential to influence CVD, stroke, blood pressure and HT. Seven blood pressure candidate genes ADD3
, and SLC14A2
were connected with cis
-lncRNA transcripts [60
]. Of these, the lncRNA NPPA-AS was selected for further investigation which revealed NPPA-AS has a demonstrated influence on the splicing of the natriuretic peptide precursor A (NPPA
) gene (present in cardiac hypertrophy and heart failure) and, therefore, has potential CVD involvement [60