2. Histone Modifications in Alzheimer’s Disease
2.1. Histone Acetylation
Histone acetylation plays a role in some crucial mechanisms, such as cognitive functions, memory and learning, response to stress, synaptic plasticity, DNA damage repair, and neuronal death
[20][23][24]. Alterations in H3K9ac were correlated with tau-associated pathology and changes in chromatin remodeling in the prefrontal cortex of AD patients compared to elderly controls
[25].
In an epigenome-wide analysis of H3K27ac conducted in the entorhinal cortex of AD patients and age-matched low-pathology controls, it was showed that H3K27ac was enriched in genes involved in Aβ and tau pathology, as well as in regions representing LOAD-associated variants
[26]. Moreover, Nativio et al. reported decreased H4K16ac across the genome of AD patients compared to controls. These results suggest that AD pathology may be way more complex, presenting distinctive mechanisms from normal aging
[23].
By combining transcriptomic, proteomic, and epigenomic analyses, a multi-omics study showed enrichment of H3K27ac and H3K9ac and a loss of H3K122ac in the temporal lobe of AD patients compared to young and elderly controls. These abnormalities were associated with the upregulation of chromatin- and transcription-associated genes, such as
CREBBP,
EP300, and
TRRAP, which encode HATs, including those responsible for the acetylation of H3K9/K27, suggesting a reconfiguration of the epigenome as a mechanism involved in AD pathology
[27].
The study of histone acetylation modifiers, such as sirtuins, the class III HDACs, has also provides valuable results for AD research and treatment.
The protein and mRNA expression levels of SIRT6 were found decreased in the brains of AD patients and a FAD mouse model. Mouse hippocampal cells treated with Aβ42 also showed decreased levels of SIRT6, along with increased acetylation levels of H3K9 and H3K56, which are SIRT6 targets. Thus, Aβ42 seems to be implicated in the decrease of SIRT6. Furthermore, SIRT6 overexpression reduced the levels of γH2AX, a marker of DNA damage, and the Aβ42-induced DNA damage. These findings provide more evidence of the role of histone acetylation and its modifier enzymes in AD
[28].
Besides SIRT6, SIRT1 has been reported to play an important protective role in AD-associated symptoms. SIRT1 deacetylates histone H1, H2, and H4 residues and other non-histone proteins, such as p53, NF-κB, and RARβ, and is often implicated in anti-inflammatory, antioxidant and anti-apoptotic responses, as well as a role in synaptic plasticity, memory and learning
[29]. Overexpression of SIRT1 is thought to reduce the levels of Aβ peptide through increasing α-secretase and, consequently, the preferential activation of the non-amyloidogenic pathway of APP cleavage. It is also reported to prevent the activation of the microglia-mediated release of pro-inflammatory factors due to Aβ toxicity
[29].
Animal model studies have also made great advances in understanding the role of histone acetylation in AD. Moreover, the development of HDACi has brought promising results for AD therapeutics.
Inhibition of HDAC3 in an AD mouse model increased histone H3 and H4 acetylation and decreased the accumulation of Aβ and tau phosphorylation while improving learning and memory in such animals
[30]. Similar results in Aβ accumulation and tau phosphorylation were also observed in cultured neurons derived from APOE ε4-carrying AD patients
[30].
Using a selective inhibitor of HDAC6, Cuadrado-Tejedor et al. observed improved memory impairment and decreased Aβ levels in the hippocampus of the Tg2576 AD mouse model and highlighted the advantages of using a specific HDACi over the pan-HDACi ones
[31].
HDACi have been used for the treatment of neurodegenerative diseases for their potential mechanisms in neuroprotection, through the upregulation of neurotrophic factors, in preventing the accumulation of neurotoxic proteins or peptides, such as Aβ, and in the downregulation of pro-inflammatory cytokines
[32][33].
The most common HDACi used for the treatment of central nervous system (CNS) diseases include vorinostat (also known as SAHA), valproic acid (VPA), trichostatin A (TSA), and sodium 4-phenylbutyrate (4-PBA), and some of them have been used to treat AD-related symptoms
[32][33].
Using VPA, Qing et al. demonstrated that this HDACi was able to reduce the production of Aβ and the formation of senile plaques while improving memory impairment in a transgenic mouse model of AD
[34]. In another study, VPA has been shown to enhance neurogenesis through the Wnt pathway and improve learning and memory abilities in the transgenic mice model for AD
[35].
Moreover, 4-PBA has also been demonstrated to reverse learning and memory deficits and decrease tau phosphorylation, besides enhancing the transcription of genes involved in synaptic plasticity through increasing histone acetylation levels in a mouse model for AD
[36]. In addition to these findings, 4-PBA was also shown to induce Aβ clearance and restore dendritic spine densities in hippocampal neurons
[37].
Treatment with the HDACi Suberoylanilide hydroxamic acid (SAHA or vorinostat) in a mouse model of AD has been observed to reverse cognitive deficits and improve memory
[38]. However, it was demonstrated that SAHA has a broader distribution on peripheral tissue and a limited effect on the brain
[39][40]. Otherwise, the combination of SAHA with other drugs seems to have a synergistic and neuroprotective effect against Aβ and tau pathology and cognitive deficits, and also reduced the levels of oxidative stress and neuroinflammatory markers while increasing the levels of CREB and neurotrophic factors, such as BDNF and GDNF
[41][42].
Finally, TSA similarly had a positive effect in reducing senile plaques and improving memory and learning behaviors in APP/PS1 mice. Such results occurred possibly due to its action towards inhibiting Aβ production or enhancing Aβ clearance
[43].
Although HDACi treatments have proved to be helpful for AD-related pathological features in mouse models or in vitro studies, there are currently no efficient established HDACi-based treatments for AD patients. However, clinical trials are being conducted and demonstrate promising results for future treatments for AD, in order to overcome the side effects and toxicity presented by the previously tested drugs
[44][45].
2.2. Histone Methylation
As well as histone acetylation, histone methylation also plays a role in important physiological mechanisms, such as regulation of transcription, alternative splicing, DNA damage responses, DNA replication, chromatin compaction, genome stability, and in a wide range of disease processes, such as cancer and neurodegenerative diseases
[46].
The role of histone methylation in AD is less understood, and studies encompassing this histone modification in AD pathophysiology have recently emerged.
An increase in H3K9me2 levels, a repressive histone modification, in the prefrontal cortex of a familial AD (FAD) mouse model was observed. The expression levels of the genes encoding the HATs that catalyze the dimethylation of H3K9,
Ehmt1, and
Ehmt2 were also increased, as well as their protein levels. Another mouse model of AD also had increased levels of H3K9me2 in the prefrontal cortex. H3K9me2 levels were also increased in the prefrontal cortex of AD patients, as well as the expression levels of
EHMT1, but not
EHMT2. In addition, the increased levels of H3K9me2 in the FAD mouse model were associated with decreased levels of the subunits of AMPA and NMDA glutamate receptors. Upon treatment with EHMT1/2 inhibitors, changes in H3K9me2 and glutamate receptors expression levels were reversed, adding more evidence for epigenetic dysregulation in AD and suggesting a therapeutic strategy targeting histone methylation for AD treatment
[47].
Wang et al. also observed an increase of H3K9me2 in cortical and hippocampal neurons of mice subjected to induced hypoxia exposure, a condition that increases Aβ production and deposition. Such increases in H3K9me2 were found in the promoter of neprilysin (
NEP), a gene that encodes one of the proteins responsible for the degradation of Aβ peptide. Thus, the downregulation of
NEP is associated with the increase of Aβ. Interestingly, the levels of G9a, an HMT that catalyzes the H3K9me2 histone mark, also increased. Knockdown of G9a was able to partially reverse the increase of H3K9me2 and prevented the decrease of
NEP [48].
Analyzing the levels of H3K4me3, a gene activation-related histone mark, Cao et al. found increased levels of this histone modification in the prefrontal cortex of both AD patients and a mouse model of tauopathy, as well as the levels of the family of HMTs that catalyze this modification. Those changes were associated with impairment of memory-related behaviors and synaptic functions, and tau hyperphosphorylation, which was recovered in the mouse model upon selective inhibition of H3K4me3 HMTs, contributing to understanding the role of histone methylation in AD pathology and providing more basis for novel treatments of AD and tauopathies
[49].
In another study, in which the levels of H3K4me3 were assessed in the CK-p25 mouse model of AD, Gjoneska et al. reported an increase in the peak enrichment of this mark in regions associated with immune response pathways, while decreased levels were observed in regions associated with synaptic and learning functions. Similar enrichment patterns were also observed in the hippocampus of AD patients
[50].
Besides identifying changes in histone acetylation marks, Nativio et al. also observed changes in a number of histone methylation marks in AD patients compared with elderly controls, with gains (H4K20me2, H3K4me2, H3K27me3, and H3K79me1) and losses (H3K79me2, H3K36me2, H4K20me3, H3K27me1, and H3K56me1) of marks associated with both gene activation and repression, thus highlighting that histone methylation dynamics may be potentially dysregulated in AD
[27].
As demonstrated in studies targeting the reversal of dysregulation of histone methylation marks by interfering in the functional role of HMTs and HDMs
[47][48], it has been shown that these histone-modifying enzymes have such an important role in this dynamic process. Thus, taking into account the role of histone methylation in memory-related functions and AD, it is reasonable to consider the maintenance of balancing between HMTs and HDMs levels for the proper functioning not only of memory but a range of processes that, once dysregulated, may trigger or contribute to the progression of AD
[51].
If increased levels of the histone methylation-modifying enzymes are able to impair memory and cognitive functions, decreased levels may also impair important functions since these proteins are involved, in addition to memory functions in the transcriptional regulation and chromatin modification pathways
[52].
Kerimoglu et al. evaluated the knockdown of the lysine methyltransferases Kmt2a and Kmt2b in hippocampal neurons of mice and observed a decrease in H3K4me3 along with impaired memory functions. However, the knockdown of these two lysine methyltransferases (KMT) impaired different gene expression regulatory pathways: genes associated with the regulation of transcription, mRNA processing, and chromatin binding were affected by the knockdown of Kmt2a and genes involved in Wnt signaling, cytokine activity, angiogenesis, and cell adhesion pathways were perturbed by the knockdown of Kmt2b. Additionally, the changes in H3K4me3 observed in neurons of mice lacking Kmt2a were similar to those found in the CK-p25 mouse model of AD neurodegeneration previously reported
[50], including decreased H3K4me3 levels in sets of genes enriched for memory- and synaptic plasticity-related categories
[53].
The role of HDM in the neurodegeneration process has also been demonstrated. Upon deletion of LSD1, a histone demethylase that demethylates specifically mono and dimethylation of H3K4 (H3K4me1/2), mutant adult mice had widespread neuronal death in the hippocampus and cortex, as well as learning and memory deficits. In addition, the transcriptional changes observed in these animals were similar to those altered in AD and frontotemporal dementia, and LSD1 is co-localized with aggregates of senile plaques and neurofibrillary tangles in AD. These results suggest a possible role of LSD1 in preventing neuronal death and consequent neurodegeneration and also reveal a mechanism of dysregulation possibly involved in AD and other neurodegenerative diseases
[54].
Thus, the important involvement of HMTs and HDMs in the regulation of histone methylation levels is clear. It is crucial in the maintenance of the appropriate levels of histone methylation marks. It is also strongly required for the proper balance between the histone methylation-modifying enzymes because either an increase or decrease in their levels may contribute to disease states, including AD. As well as for histone acetylation dynamics, research involving the administration of drugs to control suitable levels of histone methyltransferases and demethylases is emerging and exhibiting promising results for the therapeutic approach of AD
[52].
2.3. Histone Phosphorylation
Histone phosphorylation is another histone modification that has been reported to play a role in AD pathology, although few studies still address this modification in AD. This hPTM has been associated with transcriptional regulation, DNA damage repair, apoptosis, and chromatin remodeling
[55][56]. However, since it has also been associated with neuroplasticity and memory consolidation processes
[57][58], it is reasonable to consider its role in other brain functions and even in neurodegenerative diseases, including AD.
Phosphorylation is also a process required for non-histone proteins, such as TFs. Interestingly, the kinase mitogen- and stress-activated protein kinase-1 (MSK1), in addition to phosphorylate histone H3 residues, also mediates the phosphorylation and consequent activation of CREB, a TF that is a key component of the coactivator complex CREB:CBP with HAT activity and is important for transcriptional activation through histone acetylation
[59]. Thus, histone modifications are dynamic processes that can act together and are sometimes dependent on each other in a series of processes
[58].
An increase in phosphorylation of serine (S) 47 of histone H4 (H4S47p) was found in cells expressing an APP isoform and in Aβ-treated neurons. Therefore, the authors investigated if the same results were observed in brain samples of mild cognitive impairment (MCI) and AD patients. The results showed a slight increase of H4S47p in MCI and a much more significant increase in AD brain samples, demonstrating an APP and/or Aβ-mediated dysregulation in histone phosphorylation in AD
[60].
Phosphorylation has also been seen in the histone variant H2AX in brain samples of AD patients. This variant histone is phosphorylated on Ser-139 in response to DNA damage, such as double-strand break (DSB), to form γH2AX. High levels of γH2AX were found in astrocytes of the hippocampus and cortex of AD patients but not in age-matched controls, highlighting the role of astrocytes and DNA damage responses in AD, along with a better understanding of the role of histone phosphorylation in AD
[61].
Ogawa et al. also explored histone phosphorylation levels in the brain of AD patients and age-matched controls. They found an increase in phosphorylation of histone H3, specifically at Ser-10 (H3S10p), in the hippocampal tissue of AD patients. Interestingly, phosphorylated H3 was found in the cytoplasm of vulnerable neurons in AD rather than the nucleus. These findings were related to aberrant mitotic machinery and cell cycle activation, indicating a possible mechanism leading to AD neurodegeneration
[62].
In accordance with the other studies previously reported, Rao and colleagues found increased levels of histone phosphorylation in AD brains. This increase was observed in total histone H3 in the frontal cortex of AD patients compared to age-matched controls and was associated with an increase in global DNA methylation
[63].
Although there are few studies with a focus on histone phosphorylation and its role in AD pathogenesis, the studies published to date contribute to understanding how aberrant histone phosphorylation can impair neuronal and glial functions in AD and contribute to the accumulation of damage and the progressive neurodegeneration observed in the brains of AD patients.