Histone deacetylase inhibitors as potential treatment for spinal muscular atrophy

Histone acetylation plays an important role in regulation of transcription in eukaryotic cells by promoting a more relaxed chromatin structure necessary for transcriptional activation. Histone deacetylases (HDACs) remove acetyl groups and suppress gene expression. HDAC inhibitors (HDACIs) are a group of small molecules that promote gene transcription by chromatin remodeling and have been extensively studied as potential drugs for treating of spinal muscular atrophy. Various drugs in this class have been studied with regard to their efficacy in increasing the expression of survival of motor neuron (SMN) protein. In this review, we discuss the current literature on this topic and summarize the findings of the main studies in this field.


Introduction
Proximal spinal muscular atrophy (SMA) is a fatal, autosomal recessive pediatric neuromuscular disorder that is characterized by the destruction of a-motor neurons in the anterior horn of the spinal cord. SMA has an estimated incidence of 1/6,000 to 1/10,000 live births, with a carrier frequency of~1/50 individuals (Burletet al., 1996;Feldkotter et al., 2002;Kernochan et al., 2005). The criteria for classifying SMA include age of onset and disease progression, based on which SMA patients can be classified into one of four types. Entire gene deletion as well as a variety of intragenic deletions, point mutations and other truncating mutations of survival of motor neuron1 (SMN1) on chromosome 5q13that lead to loss of gene function are the cause of SMA (Clermont et al., 1994;Lefebvre et al., 1995;Burglen et al., 1996;Burlet et al., 1996). A highly related homolog of the gene, SMN2 or centromeric SMN, is retained (with a variable copy number) in all SMA patients. The substitution of a C by T at position+6 disrupts a exon splice-enhancing region in exon 7. This change results in most SMN2 transcripts lacking exon 7 and encodes a truncated protein (Feldkotter et al., 2002;Kernochan et al., 2005). SMN2 has, for many years, provided a promising opportunity for correcting SMN deficiency. The fact that SMN2 produces SMN protein, although at an insufficiently low amount, led investigators to search for ways of increasing the full-length expression of this gene in order to ensure a sufficient level of the protein. Studies in transgenic mice have shown that the insertion of eight copies of human SMN2 into the mouse genome completely rescued Smn-/mice (Smn-/-; hSMN2+/+) from the SMA phenotype (Monani et al., 2003). In humans, a high copy number of SMN2 may prevent SMN1-deficient individuals from manifesting the SMA phenotype (Prior et al., 2004). An increase in full-length SMN protein production through enhanced SMN2 expression may be achieved through promoter activation, modulation of exon 7 splicing (inclusion of exon 7 in the SMN2 transcript) or both. Another therapeutic target includes SMN1 subtle mutations. A subset of SMA patients carrying SMN1 subtle mutations is susceptible to nonsense-mediated mRNA decay (NMD) (Brichta et al., 2008). In this regard, studies aimed at identifying substances that can stabilize SMN mRNA, especially those that express the full-length protein, are of interest.
One group of drugs in particular, namely, histone deacetylase(HDAC)inhibitors, has been found to increase SMN2 promoter activity. Histone acetylation is an important epigenetic mechanism that regulates gene expression. When the N-terminus of core histones is acetylated the corresponding chromatin region is more actively transcribed because of increased accessibility to the DNA. Several drugs in this group have shown promising results in increasing SMN promoter activity as will be summarized below.
This article focuses on HDAC inhibitors that target classic HDACs and provides a comprehensive overview of current research on SMA therapy using these inhibitors. Specifically, we will discuss the characteristics and therapeutic potential of valproic acid, phenylbutyrate, benzamide M344, suberoylanilidehydroxamic acid, LBH589, trichostatin A, MS-275, romidepsin, resveratrol, curcumin and epigallocathecin gallate.

HDACs and HDAC inhibitors
Histone remodeling by acetylation and/or deacetylation plays an important role in the transcriptional regulation of eukaryotic cells. Histone acetylation produces a more relaxed chromatin structure that allows transcriptional activation (Kernochan et al., 2005;Riester et al., 2007). This is achieved through the acetylation of lysine residues that imparts a negative charge to the affected amino acid which in turn relaxes the chromatin. In this regard, HDACs are actually "lysinedeacetylases" (Grayson et al., 2010;Xu et al., 2007). HDACs therefore repress transcription through histone deacetylation.
HDACs form a large family of enzymes and have been classified into two groups based on their co-enzyme requirements and sequence similarity to yeast HDACs. These two groups, known as classic HDACs and Sir2related HDACs (Sirtuins or Class III HDACs), are activated by Zn 2+ and NAD + , respectively. Classic HDACs are subdivided into three smaller classes that include HDAC-I (Ia, Ib and Ic), HDAC-II (IIa and IIb) and HDAC-IV. Each of these smaller classes consists of functional HDAC enzymes (HDAC1 to HDAC11) that are targeted by different HDAC inhibitors (Table 1A,B).Overall, there are 11 classic HDAC enzymes while the Sirtuins contain seven members (Sirt1-Sirt7) (Xu et al., 2007;Nakagawa and Guarente, 2011).
HDAC inhibitors selectively alter gene transcription through chromatin remodeling and by changing the protein structure of transcription factor complexes (Kernochan et al., 2005;Riester et al., 2007). HDAC inhibitors generally consist of three domains: a linker region, a capping group and a metal moeity (Dayangac-Erden et al., 2011).

Valproic acid
Valproic acid (VPA) or Depakene is a Federal Drug Administration (FDA)-approved drug with a terminal halflife (t 1/2 ) of 8-10 h in human serum and is frequently used to treat epilepsy, mood disorders and migraine (Brichta et al., 2003). Although VPA is associated with few neurological side effects, hematological and hepatic side effects are well known (Cotariu and Zaidman, 1988;Lackmann, 2004;Tong et al., 2005). VPA increases SMN protein levels through transcriptional activation but also increases the expression of additional serine/arginine (SR)-rich proteins that may have important implications for disorders (including SMA) caused by mutations that result in alternative splicing. While promising results have been obtained in-vitro, clinical trials have yielded variable results (Table 2).
Chemical characteristics: VPA is a simple eightcarbon branched fatty acid (carboxylic acid;C 8 H 14 O 2 ) designated as 2-propylpentanoic acid but is also known as dipropylacetic acid.

Phenylbutyrate
Phenyl butyric acid (PBA) or buphenyl is a shortchain fatty acid that has been clinically tested as an anticancer drug. In normal tissues, PBA shows little toxicity 300 Mohseni et al.  No adverse effects observed and provides protection against various stimuli. Sodium PBA is a pro-drug that is rapidly metabolized to phenylacetate, a metabolically-active derivative. Phenylacetate conjugates with glutamine via acetylation to form phenylacetylglutamine that is excreted by the kidneys. PBA shows anticancer activity that is generally attributed to its activity as an HDAC inhibitor. Table 3 summarizes studies that have investigated PBA in SMA.
Chemical characteristics: PBA (molecular weight: 186; C 10 H 11 O 2 Na) is known chemically as 4-phenylbutyric acid and is usually supplied as a sodium salt.

Benzamide M344
M344 is a HDAC inhibitor that increases the level of hyperacetylated histone H4 and significantly increases SMN2 mRNA/protein levels in SMA cells by inducing ter-minal cell differentiation. M344 shows a three-fold selectivity for inhibition of HDAC6 over HDAC1. Table 4 summarizes studies that have investigated benzamide M344 in SMA.

LBH589
LBH589 (Panobinostat) is a potent putative anticancer drug in numerous cancer cell lines and was given orphan drug status for the treatment of cutaneous T-cell lymphoma (CTCL) by the FDA in 2007. LBH589 is also a novel hydroxamic-acid-derived HDAC inhibitor that is active against all classes of HDACs at low nanomolar con- 302 Mohseni et al.

Suberoylanilidehydroxamic acid (SAHA)
Suberoylanilidehydroxamic acid (SAHA;zolinza or vorinostat) was initially approved for the treatment of cutaneous T-cell lymphoma (CTCL). Vorinostat, an FDAapproved pan-histone deacetylase inhibitor, is a potentially useful drug for clinical trials in SMA patients. Some of this drugs side-effect includes gastrointestinal symptoms, constitutional symptoms (thrombocytopenia, anemia), taste disorders, pulmonary embolism and anemia. Severe thrombocytopenia and gastrointestinal bleeding have been reported with the concomitant use of zolinza and other HDAC inhibitors, e.g.,valproic acid. Table 6 summarizes studies that have investigated SAHA in SMA.
Chemical characteristics: SAHA (N-hydroxy-N'phenyloctanediamide; C 14 H 20 N 2 O 3 ) is poorly soluble in water, slightly soluble in ethanol, isopropanol and acetone, freely soluble in dimethyl sulfoxide and insoluble in methylene chloride.

Trichostatin A (TSA)
Trichostatin A (TSA), originally developed as an antifungal drug, is a member of a large class of HDAC inhibitors that has a broad spectrum of epigenetic activities. TSA selectively inhibits class I and II mammalian HDAC. TSA alters gene expression by interfering with the removal of acetyl groups from histones by HDAC and therefore alters the ability of DNA transcription factors to access the DNA within chromatin. TSA is harmful by inhalation and is irritating to the eyes, respiratory system and skin. Table 7 summarizes the studies on TSA in SMA.
Histone deacetylase inhibitors for spinal muscular atrophy 303 No cytotoxic effects at up to 500 nM USA Ex vivo TSA improved short-term function and produced long-lasting stabilization of the SMA motor unit. In affected mice treated with TSA and a dietary supplementation the median survival time increased by up to 38 days (170%) as compared to non-treated mice.
Chemical characteristics: The molecular formula of Entinostat is C 21 H 20 N 4 O 3 .

Romidepsin
Romidepsin (Istodex or FK228), an HDAC inhibitor from Chromobacterium violaceum, is a bicyclic depsipeptide. Romidepsin is indicated for the treatment of CTCL in patients who have received at least one prior systemic therapy. Romidepsin shows hematologic and non-hematologic toxicity at high doses. Table 9 summarizes a study that investigated the usefulness of romidepsinin SMA.

Resveratrol
Resveratrol (Kojo-Kon, Phytoalexin, Phytoestrogen and SRT-501) is a chemical found in red wine, red grape skins, purple grape juice, mulberries and in smaller amounts in peanuts. Resveratrol is used against hardening of the arteries (atherosclerosis), high cholesterol and for the prevention of cancer. Resveratrol may increase the risk of bleeding. Table 10 summarizes studies that have investigated resveratrol in SMA.

Curcumin
Curcumin is a mixture of compounds derived from the curry spice turmeric and is used as an herbal supplement. Curcumin (diferuloylmethane) is a new HDAC inhibitor that inhibits the expression of class I HDACs (HDAC1, HDAC3and HDAC8). Curcumin possesses a spectrum of pharmacological properties that have been attributed primarily to its inhibition of metabolic enzymes. Curcumin has been alleged to have antioxidant, antiviral, anti-inflammatory and anticancer activities, as well as cholesterol-lowering effects.
Chemical characteristics: Curcumin, a natural polyphenol and the major component of turmeric has the molecular formula C 21 H 20 O 6 .

Epigallocatechin gallate
Epigallocatechin gallate (EGCG; Sinecatechins or Veregen), a partially purified fraction obtained from a water extract of green tea (Camellia sinensis) leaves, is used topically and is a potent antioxidant.

Discussion
Eight of the 11 known HDACs were inhibited by the compounds reviewed here; HDAC4, HDAC7 and HDAC10 were not inhibited by any of the compounds. As shown in Table 1B, the fold increase of full-length SMN2 transcripts or SMN protein varied considerably (from 0.4 to 10).
Five compounds (VPA, M344, resveratrol, EGCG and curcumin) acted by two mechanisms, namely, (1) by increasing the overall SMN2 expression through inhibition 304 Mohseni et al.  (2) by increasing the incorporation of exon 7 into the SMN2 transcripts through the activation of splicing factors. However, the latter three compounds induced only a minimal increase in the total SMN2 transcript level. Nevertheless, these compounds may still have useful chemical properties because they are derived from natural products and show few or no adverse effects. In this regard, insilico analyses may be helpful in optimizing the design of molecules with greater effect on SMN2 while retaining their safety.
In addition to HDAC inhibition, an increase in the overall SMN2 transcript level can also be achieved by de-methylation of the SMN2 gene. An increase in SMN2 expression through de-methylation, i.e., bypassing SMN2 gene silencing, was recently suggested for SAHA, MS275 and Romidepsin (Haukeet al., 2009), and indicated that these three drugs to have a double mechanism of action in addition to inhibiting targeted HDACs. However, demethylation contributed to only 5% of the total increase in full-length transcripts.
In contrast, inhibition of HDAC6 by LBH-589 and M344 resulted in the highest fold increase of full-length transcripts, even when compared to inhibition of multiple HDACs. Li et al. (2013) indicated that, unlike other deacetylases, HDAC6 has a unique substrate specificity for non-histone proteins. This diversity of functions for HDAC6 suggests that this enzyme could be a potential therapeutic target for the treatment of a wide range of diseases. In this regard, finding an inhibitor of HDAC6 may help in the search for a potent SMN2 expression activator. It would also be worthwhile to study the effects of currently known HDAC6 inhibitors in SMA cell lines. Once the structure of HDAC6 is known molecular docking strategies may be used to identify natural or synthetic inhibitors of this enzyme.
Only two of the HDAC inhibitors discussed here (PBA and VPA) have entered clinical trials for human use. The results of these clinical trials have varied considerably and a systematic review of potential drugs for treating SMA found that none of them, including HDAC inhibitors, were efficacious in treating this condition (Wadman et al., 2012a,b).

Conclusion
We have summarized various studies that have examined the usefulness of HDAC inhibitors for treating SMA. Naturally-derived HDAC inhibitors (also summarized here) are less toxic but also show less therapeutic promise. Given the therapeutic potential of HDAC inhibitors and their theoretical mechanism of action, a search for further inhibitors is warranted in an effort to identify molecules with suitable properties (high blood-brain barrier penetration and minimal/tolerable adverse effects) that can be used to correct the molecular pathology of SMA. Turkey In vitro (cell-based) Curcumin increased FL-SMN mRNA level significantly by up to 1.7 fold and caused a concentration-dependent increase in exon 7 inclusions.  Sakla and Lorson (2008) USA In vitro (cell-based) Resveratrol elevated SMN2-luciferase expression by six fold and increased the exon 7-inclusion by 1.4 fold in a luciferase assay. These effects translated into only a one-fold increase in the full-length SMN2 transcript level.

No Reported
Not reported Dayangac-Erden et al. (2009) Turkey In vitro (cell-based) Resveratrol increased the full-length SMN2 mRNA and protein levels by 1.3-fold.
Not reported