Histone Deacetylase Inhibitors for Treating a Spectrum of Diseases Not Related to Cancer

Charles A Dinarello,1,2 Gianluca Fossati,3 and Paolo Mascagni3

1Department of Medicine, University of Colorado Denver, Aurora, Colorado, United States of America; the 2Department of Medicine, Radboud University Nijmegen Medical Center, Nijmegen, the Netherlands; and the 3Centre for Research, Italfarmaco, S.p.A., Cinisello Balsamo, Italy

This issue of Molecular Medicine contains 14 original research reports and state-of-the-art reviews on histone deacetylase in- hibitors (HDACi’s), which are being studied in models of a broad range of diseases not related to the proapoptotic properties used to treat cancer. The spectrum of these diseases responsive to HDACi’s is for the most part due to several antiinflammatory properties, often observed in vitro but importantly also in animal models. One unifying property is a reduction in cytokine pro- duction as well as inhibition of cytokine postreceptor signaling. Distinct from their use in cancer, the reduction in inflammation by HDACi’s is consistently observed at low concentrations compared with the higher concentrations required for killing tumor cells. This characteristic makes HDACi’s attractive candidates for treating chronic diseases, since low doses are well tolerated. For ex- ample, low oral doses of the HDACi givinostat have been used in children to reduce arthritis and are well tolerated. In addition to the antiinflammatory properties, HDACi’s have shown promise in models of neurodegenerative disorders, and HDACi’s also hold promise to drive HIV-1 out of latently infected cells. No one molecular mechanism accounts for the non–cancer-related proper- ties of HDACi’s, since there are 18 genes coding for histone deacetylases. Rather, there are mechanisms unique for the patho- logical process of specific cell types. In this overview, we summarize the preclinical data on HDACi’s for therapy in a wide spec- trum of diseases unrelated to the treatment of cancer. The data suggest the use of HDACi’s in treating autoimmune as well as chronic inflammatory diseases. © 2011 The Feinstein Institute for Medical Research, www.feinsteininstitute.org Online address: http://www.molmed.org doi: 10.2119/molmed.2011.00116

INTRODUCTION about the death of malignant cells in vitro As cytokine-driven inflammation plays In this issue of Molecular Medicine, we and in tumor-bearing mice. This charac- a fundamental role in nearly all acute as assembled a series of articles on the use teristic of an effective low dose suggests well as chronic diseases, targeting cy- of histone deacetylase (HDAC) inhibitors that the traditional mechanism of action tokines in metabolic, neurodegenerative, (HDACi’s) in the treatment of a broad of HDACi’s (that is, hyperacetylation of cardiovascular, bowel and joint diseases spectrum of diseases not related to can- nuclear histones with increased expres- has become a major area of investigation cer. Presently, the spectrum of diseases sion of proapoptotic genes) may not ac- (1). The importance of cytokines in the being evaluated are mostly inflammatory count for the antiinflammatory proper- pathogenesis of these diseases has been or autoimmune in nature. However, ties of HDACi’s. Acetylation of validated by the widespread use of anti- orally active drugs that are safe and anti- cytoplasmic proteins may explain the cytokine monoclonal antibodies and now inflammatory are ideal for treating unique properties of low doses of includes Type 1 and 2 diabetes, as re- chronic degenerative diseases (1). The in HDACi’s. In this issue of Molecular Medi- viewed by Mandrup-Poulsen et al. (2), vitro and in vivo studies of HDACi’s in cine, each contribution considers the Donath and Shoelson (3), and Dinarello models of inflammatory and autoim- amelioration of disease severity in a spe- et al. (4). The attractive aspect of mune diseases have required doses con- cific pathological model and proposes HDACi’s is that they are orally active, siderably lower than the concentrations the likely mechanism of action of and low concentrations are most effec- of HDACi’s that are required to bring HDACi’s. tive in reducing inflammation in humans (5) and animal models (6). The success of HDACi’s in the treat- Address correspondence and reprint requests to Charles A Dinarello, Department of Medi- ment of inflammatory diseases will de- cine, Division of Infectious Diseases, University of Colorado Denver, 12700 East 19th Avenue, pend on two factors: lack of organ toxic- Aurora, CO 80045. Phone: 303-724-6172; Fax: 303-724-6178; E-mail: [email protected]. ity and tolerability as well as the Submitted March 29, 2011; Accepted for publication May 4, 2011; Epub specificity of the inhibitor for the rele- (www.molmed.org) ahead of print May 5, 2011. vant HDAC in a particular pathological

MOL MED 17(5-6)333-352, MAY-JUNE 2011 | DINARELLO ET AL. | 333 HDAC INHIBITORS IN NONCANCER DISEASES

process. Indeed, the future development With hyperacetylation of nuclear his- fective in reducing disease severity in an- of HDACi’s should be focused on selec- tones, chromatin unravels and transcrip- imal models of inflammation and im- tive inhibitors, since there are 18 distinct tion factors can now bind to DNA and mune disorders are due to acetylation of HDACs (7). It is also expected that inhi- initiate the synthesis of RNA coding for cytoplasmic proteins and not nuclear his- bition of specific HDACs will offer opti- proapoptotic genes. For the most part, tones, although hyperacetylation of both mal efficacy depending on the dominant the doses of vorinostat and romidepsin likely take place at the same time. Com- cell type in a particular disease, for ex- used to treat humans with cancer are plexes of HDAC with histones as well as ample, reducing interferon (IFN)-γ pro- based on maximal tolerated doses. It is cytoplasmic proteins in the same cell duction by a specific HDAC in T cells. assumed that plasma levels reach those challenges the concept that targeting spe- For inflammatory diseases such as gouty concentrations observed for increased ex- cific HDACs with small-molecule arthritis or Type 2 diabetes, inhibition of pression of proapoptotic genes in tumor HDACi’s may not result in the same ef- interleukin (IL)-1β production or secre- cells in vitro. Knowing the concentration fect or results in the same therapeutic tion would best use an HDACi that regu- of a particular HDACi for inducing advantage. lates caspase-1. apoptotic cell death in primary tumor In general, HDACs are nonredundant. cells in vitro is often not possible, and the A knockdown of class I HDACs with HISTORICAL BACKGROUND success of any HDACi in the treatment siRNA resulted in cellular changes simi- HDACi’s were initially studied for of humans with cancer is thus measured lar to those observed with small-mole- their ability to increase gene expression. by a reduction in the tumor burden. cule HDACi’s (12). In mice deficient in Phenylbutyrate was used for many However, in terms of the antiinflamma- HDAC-6, there is hyperacetylation of years to increase gene expression of tory properties of HDACi’s, the in- HDAC-6, but otherwise the mice are fetal hemoglobin in patients with a ge- hibitors are studied in primary cells in healthy. Mice with null mutations for netic basis for anemia. Phenylbutyrate vitro. As stated, the most distinguishing HDAC-1 do not progress beyond embry- also has an excellent safety record. Oral property of the antiinflammatory proper- onic stages (13), whereas mice deficient phenylbutyrate was used in children ties of HDACi’s is that significantly in HDAC-2 succumb to an early death and adults with sickle cell disease. De- lower concentrations are used in models because of hyperproliferation of car- spite the increment in hemoglobin F of autoimmune diseases compared with diomyocytes (13). For example, mice de- and reticulocytes, the use of 30–40 cap- those used to reduce tumors in mice. A ficient in HDAC-2 die with cardiomy- sules a day is not an optimal therapy similar example exists for the inhibition opathy. HDAC-8-deficient mice die soon (8,9). Today, however, the increasing of cytokines in vitro. For example, the after birth because of malformations in number of orally active, synthetic 50% inhibition of proliferation of A549 calvarial and facial bones (14). A null HDACi’s are primarily developed to tumor cells requires 5 μmol/L, whereas mutation in HDAC-3 in the liver dis- treat cancer. The development of to inhibit 50% of IL-1β secretion requires rupts normal metabolic homeostasis (15). HDACi’s for treatment of cancer is 50 nmol/L in freshly obtained human The deletion of HDAC-8 as with other based on de-repression of genes that peripheral blood mononuclear cells HDAC deletion studies suggests that participate in endogenous proapoptotic (PBMCs) (10), and similar data have HDAC functions to suppress either the pathways and bring about a selective been reported for ITF2357 (11). Thus, ability of particular transcription factors death of malignant cells while sparing when considering the efficacy of an active during embryonic development or healthy cells. By use of this mechanism, HDACi to treat a cytokine-driven inflam- the expression and synthesis of other HDACi’s would avoid the toxic effects matory disease, it is possible to draw transcription factors, which serve to sup- of many chemotherapeutic drugs. It some conclusions that the in vivo efficacy press genes. Silencing of HDAC-1, was Paul Marks and his coworkers that is related to an in vitro concentration and, HDAC-2 or HDAC-3 by using siRNA de- brought the first HDACi to be ap- because of the lower concentrations, creased IFNγ-inducible genes, whereas proved, suberoylanilide hydroxamic more likely to be achieved in vivo. overexpression of these same HDACs in- acid (SAHA). SAHA (generically As discussed in the various reports in creased signal transducers and activators vorinostat) and romidepsin (FK228) are this issue, HDACi’s also hyperacetylate of transcription (STAT)-1-dependent approved for the treatment of cuta- nonhistones such as cytosolic proteins, genes (16). neous T-cell lymphoma (CTCL). Similar including transcription factors them- to other HDACi’s, SAHA and FK228 in- selves. Because HDACs deacetylate the HDACi’s REDUCE THE PRODUCTION OF crease several genes that induce apo- highly conserved N-terminal lysines PROINFLAMMATORY CYTOKINES ptosis in malignant cells and are consis- present on histones, it is also appropriate For reducing disease severity in sev- tent with the well-studied mechanism to use KDACi’s for lysine deacetylase in- eral immune and inflammatory condi- of all HDACi’s (that is, hyperacetyla- hibitors. In fact, it is likely that some tions, synthetic glucocorticoids remain tion of nuclear histones). mechanisms by which HDACi’s are ef- the mainstay of therapies. However, with

334 | DINARELLO ET AL. | MOL MED 17(5-6)333-352, MAY-JUNE 2011 REVIEW ARTICLE

Table 1. Properties of HDACi’s for treating disease not related to cancer. specific antibodies to cytokines such as tumor necrosis factor (TNF)-α, IL-1β, Systemic inflammation IL-6 receptors and others, anticytokine- Reviewed in (79) based therapies have found a place in the Improved survival after hemorrhage (89) treatment of autoimmune diseases such Decreased cytokinemia after systemic endotoxin (10,11,136) as rheumatoid arthritis, inflammatory Decreased NO production (6,10,32) bowel disease, psoriasis and several oth- Protection against hepatitis after intravenous concanavalin A (11) ers. Reducing the activity of IL-1β has Suppression of vascular cell adhesion molecule-1 expression (137) had a major impact on the treatment of In vitro cytokine production autoinflammatory diseases (17). Without Decreased gene expression and synthesis of LPS-induced TNFα and IFNγ in human question, neutralization of specific cy- peripheral blood mononuclear cells (10,11,138) tokines has canonized their causative Decreased IL-1β-induced IL-6 in synovial cells and human blood monocytes (23) role in inflammation and has changed Reduced IFNγ in mouse trophoblasts (139) the lives of millions of patients with Decreased production of IL-6 in PBMCs stimulated with IL-12 plus IL-18 (11) β these diseases (1). Compared with the Decreased secretion of IL-1 (10,11,52) metabolic consequences of long-term Decreased IL-12 and IL-17 production from monocytes and dendritic cells (10,20,50) Decreased IL-6 from stromal cells (104) glucocorticoids, anticytokine therapies Reduced IL-1β-induced PGE2 and NO in human chrondrocytes (140) are a major improvement in treating an Decreased ischemia-induced retinal TNFα production (141) increasing number of diseases, but par- Downregulation of IL-12 transcription in transformed lung epithelial cells (142) enteral administration is a major draw- Increased PGD2 production and suppression of matrix metalloproteinases (82) back of anticytokine therapy, and de- Suppression of c-jun and transcription of COX-2 (83) creased host defense against infection is Models of inflammatory bowel disease another. Nevertheless, anticytokine ther- Reviewed in (74) apy is nearly devoid of organ toxicity. Reduced disease severity in dextran sodium sulfate and trinitrobenzene sulfonic acid Because inflammation plays a fundamen- (TNBS) colitis (143,144) tal role in nearly all chronic inflamma- Decreased cytokine levels in colitis model (143,144) tory and degenerative diseases, in- Reduced severity of colitis in HDAC-9-deficient mice (145) hibitors of proinflammatory cytokine Suppression of COX-2 activation in colon cells (146) production as well as activities has be- Immunosuppressive properties come a major area of investigation for Reviewed in (43,75,76) orally active drugs. As summarized in Survival benefit in GvHD (71,72,147) Table 1, orally active HDACi’s are effec- Sparing effect on graft versus leukemia (71) in vitro + tive in reducing cytokines and in Increased Foxp3 T-regulatory cells (148,149) various animal models. Reduced incidence of diabetes in NOD mouse (36) SAHA was developed to treat cancer, Induction of antigen-specific anergy in lymphocytes (150) but in studies published in 2002, we Improved allograft transplantation (148,151) added SAHA to freshly obtained human Induction of IDO and inhibition of dendritic cell maturation (44) peripheral blood mononuclear cells Reduced nephritis in lupus-prone mice (119,120) (PBMCs) stimulated with lipopolysac- Decrease disease severity in experimental allergic encephalitis (76,152) charide (LPS) (10). There was 50% reduc- Suppression of Th1 polarization of murine dendritic cells (153) tion in TNFα, IL-1β, IL-12 and IFNγ at Inhibition of IL-2 gene expression in T cells (154,155) 100–200 nmol/L, a 20-fold lower concen- Reduced cytokine production in primary human T cells stimulated with anti-CD3/CD28 (156) tration that that required for inhibition of Inhibition of CD154 (CD40L) expression in T cells (157) tumor cell proliferation in vitro. When Models of arthritis stimulated with the combination of IL-12 Reviewed in (25,26) plus IL-18, IFNγ was reduced by 85%. Reduced joint destruction in collagen-induced arthritis (25,158,159) Steady-state mRNA levels for LPS- Decreased bone and cartilage loss (25) induced TNFα and IFNγ were markedly Lower cytokine and chemokine levels (23,24) decreased, whereas levels for IL-8 and Decreased synovial cell proliferation (160) IL-1β mRNA were unaffected. Although Abrogation of TGFβ-1-induced fibroblast-myofibroblast differentiation (161) β the levels of the IL-1β precursor were not Inhibition of IL-1 -induced matrix metalloproteinase expression in human articular decreased, secretion of IL-1β was re- chondrocytes (162) duced at 100–200 nmol/L. Continued

MOL MED 17(5-6)333-352, MAY-JUNE 2011 | DINARELLO ET AL. | 335 HDAC INHIBITORS IN NONCANCER DISEASES

Table 1. Continued. CCR5 expression on CD4+ T cells and Heart failure monocytes (19). Reviewed in (63) In related studies, human PBMCs Reduced Hop-mediated hypertrophy and heart failure (163) were differentiated into dendritic cells, Decreased left ventricular cardiac hypertrophy (164) and the effect of the HDACi tricho- Reduced cardiac arrhythmias in dystrophic mice (165) statin A (TSA) as well as SAHA was in- Brain and neurological systems vestigated on the production of IL-12 Reviewed in (77) and (76) and IL-17. There was a reduction of 86% Improved neurological recovery after closed head trauma (93) and 83%, respectively, by SAHA. In Reduced gliosis and neuronal apoptosis (93,166) these differentiated dendritic cells, the Decreased neuroinflammation in glial cells (41) T-cell chemokines CXCL9, -10 and -11 Reduced brain infarct after cerebral artery occlusion (115,167,168) were also reduced (20). Oral adminis- Decreased disease severity in models of Huntington disease (96,97,99) tration of 1.0–10 mg/kg ITF2357 to Reduced loss of function in ALS model (169,170) mice reduced LPS-induced serum Effects in multiple sclerosis (76,152) TNFα and IFNγ by >50%. However, Models of diabetes anti–CD3-induced cytokines were not Reviewed in (34) suppressed by ITF2357 in PBMCs ei- Decreased death in IL-1β-induced pancreatic insulin-producing β cells (6,32-35,171) ther in vitro or in the circulation in Decreased death in IL-1β-induced insulin-producing INS cells (32,34,35) mice. In concanavalin-A–induced hepa- Protection from streptozotocin-induced diabetes (6) titis, 1 mg/kg oral ITF2357 significantly Decreased islet cytokine production (6) reduced liver damage at the same level β Inhibition of IL-1 -induced NO production by pancreatic islets (6,32,34,35) as 5 mg/kg. In mice subjected to Fibrotic diseases streptozotocin-induced diabetes, an oral Prevention of renal interstitial fibrosis after ureteral obstruction (172) dose of 2.5 mg/kg provided protection Decreased left ventricular cardiac hypertrophy (164) (6). Thus, low, nonapoptotic concentra- Increased basic morphogenic protein 7 in the regenerative response to renal ischemia (173) tions of the HDAC inhibitor ITF2357 re- Reduced fibrotic changes in tubulointerstitial injury (174) duce proinflammatory cytokine produc- Inhibition of hypoxia-induced angiogenesis (175) tion in primary cells in vitro and exhibit β Decreased TGF -1-induced renal injury (176) antiinflammatory effects in vivo. In con- Suppression of epithelial-to-mesenchymal transition induced by TGFβ (177) trast, treatment of murine macrophages Muscular dystrophy with the histone deacetylase inhibitor Reviewed in (78) LAQ824 induced chromatin changes in HIV-1 purging the IL-10 gene promoter that results in Reviewed in (38) recruitment of the transcriptional re- In vitro studies (19,38,132,133) pressor HDAC11 (21). This result dimin- Studies in humans (116,131) ishes IL-10 production and induction of inflammatory cells. These results are in- consistent with the antiinflammatory A single oral dose of SAHA to mice be- duced by 50% at 12.5–25 nmol/L. There properties of HDACi’s. fore LPS dose-dependently reduced cir- was no evidence of cell death in LPS- Panobalinostat (LBH589) was tested for culating TNFα, IL-1β, IL-6 and IFNγ. The stimulated PBMCs at 100 nmol/L treatment of multiple myeloma and lym- effect of HDACi’s on cytokine production ITF2357, by using assays for DNA degra- phomas. At clinically relevant concentra- was also studied with ITF2357 (generic dation, annexin V and caspase-3/7. tions added to cultured human dendritic givinostat) (11). In PBMCs, ITF2357 re- There was a 50% to 90% reduction in cells (DC), LBH589 reduced the surface duced by 50% the release of TNFα at con- LPS-induced steady-state levels of TNFα molecule expression on immature and centrations of 10–20 nmol/L, the intracel- and IFNγ but no reduction of IL-1β or mature DCs, which was associated with lular levels of IL-1α at 12 nmol/L, the IL-8 (11). Similar to SAHA, the secretion DC maturation, antigen presentation and secretion of IL-1β at 12.5–25 nmol/L and of IL-1β was reduced, suggesting that T-cell costimulation (22). LBH589 also sig- the production of IFNγ at 25 nmol/L. ITF2357 affects caspase-1 and proteins nificantly reduced the production of Similar to SAHA (10), there was no re- that are required for activation of TNFα, IL-6, IL-10, IL-12 and IL-23 stimu- duction in IL-8 in these same cultures. By caspase-1 (18). Although neither SAHA lated with LPS. It was also reported that using the combination of IL-12 plus nor ITF2357 reduced the chemokine IL-8, the RelB component of nuclear factor IL-18, IFNγ and IL-6 production was re- ITF2357 did reduce surface CXCR4 and (NF)-κB had a critical role in the antiin-

336 | DINARELLO ET AL. | MOL MED 17(5-6)333-352, MAY-JUNE 2011 REVIEW ARTICLE

flammatory and immunosuppressive marily in myeloid lineage cells such as particularly cardiovascular disease. For mechanism of this HDACi (22). monocytes and macrophages. These cells example, safe and specific antiinflamma- Because there are 18 separate second are, without doubt, major players in in- tory agents are sought for the prevention coding for HDAC, the specificity of an flammation because of infections; but, of cytokine-induced destruction of pan- HDACi for a particular HDAC or a clus- most human disease is due to sterile in- creatic islet β cells. The immune- ter of related HDACs likely affects its flammation. In the models of sterile in- mediated elimination of pancreatic mechanism of action for reducing inflam- flammation, a hypoxic event is often the β cells in Type 1 diabetes involves release mation via inhibition of cytokines. trigger and, with hypoxia, there is cell of cytotoxic cytokines such as IL-1β and ITF2357, SAHA and generally all death and the release of cell contents. As IFNγ, which induce β-cell death in vitro hydroxamate-based HDACi’s do not in- such, cytokines and other intracellular by mechanisms that are both dependent hibit sirtuins, 7 of the 18 known HDACs. components initiate the inflammatory and independent of NO (2). SAHA (10) Another consideration is the cell target trigger. For example, IL-1α and HMGB1 inhibits IL-1β-induced NO in mouse being affected by HDAC inhibition. Al- are inflammatory molecules that are re- macrophages and also in rat primary though it can be assumed that all leased from dead cells and are highly in- islet cells and the INS insulinoma cell HDACi’s enter the cell by a simple con- flammatory (27–29). In autoimmune dis- line (32). SAHA reduced the cytokine- centration gradient in that there are no eases, the trigger for inflammation is not mediated decrease in insulin secretion specialized membrane channels required via the TLR pathway but rather via cy- and increase in iNOS levels, NO forma- for entry, the antiinflammatory mecha- tokines such as IL-17 (30). In Type 1 dia- tion and apoptosis (32). A similar effect nism is cell specific for the disease model. betes, IL-1β is highly toxic, inducing cell of ITF2357 was observed on IL-1β- For example, in rheumatoid arthritis the death of the insulin-producing β cells as stimulated mouse and rat primary islets fibroblast-like synoviocyte plays an im- well as nitric oxide (NO) (2). Similarly, and the INS cell line (6). Although there portant role in the pathological processes IL-1β is toxic for the β cell in Type 2 dia- was no effect of SAHA on IL-1β- of joint destruction. These cells produce betes (3). In bone loss, IL-1β and TNFα mediated degradation of IκB or binding several cytokines, both inflammatory as are the stimulators of osteoclast activa- of NFκB to DNA, SAHA reduced the sec- well as antiinflammatory, chemokines tion. Therefore, suppression of cytokine ond phase of NFκB phosphorylation (32). and prostaglandins. When stimulated signaling may be the more relevant anti- As reviewed in detail, HDAC inhibition with IL-1β, the fibroblast-like synoviocyte inflammatory mechanism for HDACi’s prevents cytokine-induced β-cell apopto- produces copious amounts of IL-6, a rather than TLR-induced cytokines. sis and impaired β-cell function (34). known growth factor for the invasive In vivo, during streptozotocin-induced pannus of joints in rheumatoid arthritis. Diabetes β-cell destruction, ITF2357 protected the HDACi’s reduce IL-1β-driven IL-6 pro- Because targeting IL-1β-mediated in- insulin-producing β cell following an oral duction as well as reduce synovial cell flammation protects islets in human trials dose of 1.25–2.5 mg/kg (6). Serum NO survival (23). The mechanism for the re- (31,32), the use of oral HDACi’s to target levels returned to nondiabetic values, duction in IL-6 production is an acceler- islet inflammation should be considered islet function improved and glucose ated degradation of IL-6 mRNA rather for both Type 1 and Type 2 diabetes. In clearance increased. In vitro, at 25 and 250 than inhibition of components of NFκB vitro HDACi’s reduced cytokine- nmol/L, ITF2357 increased islet cell via- signaling (24). As reported in this issue, mediated NO formation and the decline bility, enhanced insulin secretion, inhib- HDACi’s suppress IL-1β-driven loss of in insulin secretion in isolated rat islets ited chemokine production and reduced bone and cartilage (25), and the targets (32,33). Whereas the initial studies on the IL-1β-induced iNOS levels. In peritoneal are the chondrocyte and osteoclast, since antiinflammatory effects of HDACi’s fo- macrophages and splenocytes, ITF2357 these are key pathways for destructive cused on a reduction in the production of inhibited the production of NO, as well joint disease in both rheumatoid arthritis IL-1β, TNFα, IFNγ and other proinflam- as that of TNFα and IFNγ at a 50% (me- as well as osteoarthritis (26). matory cytokines from LPS-stimulated dian) inhibiting concentration IC50 of freshly obtained human PBMCs, other 25–50 nmol/L (6). Thus, at clinically rele- HDACi’s FOR INHIBITION OF CYTOKINE investigators examined the ability of vant doses, the orally active HDAC in- SIGNALING HDACi’s to inhibit cytokine-induced ef- hibitor ITF2357 favors β-cell survival dur- Many studies on HDACi’s focus on in- fects. If HDACi’s are to be used to treat a ing inflammatory conditions. hibition of LPS-induced cytokine produc- broad spectrum of inflammatory as well The nonobese diabetic (NOD) mouse is tion. Some of these reports are listed in as autoimmune diseases, cytokine-driven often used as the model for Type 1 dia- Table 1. Although the LPS pathway as inflammation is more relevant than in- betes in humans. Because the HDACi’s well as all Toll-like receptor (TLR) ago- flammation driven by microbial prod- SAHA, ITF2357 or TSA protect the nists are commonly studied with ucts. Indeed, sterile inflammation is cer- insulin-producing pancreatic β cell in HDACi’s, the signaling pathway is pri- tainly the dominant cause of disease, primary rat islets (6,32,33,35), in primary

MOL MED 17(5-6)333-352, MAY-JUNE 2011 | DINARELLO ET AL. | 337 HDAC INHIBITORS IN NONCANCER DISEASES

mouse islets (6) or the rat insulinoma cell was also a marked decreased expression stimulation. For example, IL-1β-induced line (6,32,35), it was not unexpected that of CCR5 on CD14-positive blood mono- phosphorylation of IκB is reduced by inhibition of HDACs by TSA adminis- cytes (38). Others have reported that SAHA, suggesting that the antiinflam- tered once a week was reported for re- HDACi’s decrease CXCR4 protein and matory properties of HDACi’s may by ducing the incidence of diabetes that mRNA levels in leukemia cell lines and due to reduced NFκB activity (32,35). As consistently develops in these mice (36). lymphoblasts from patients with child- discussed below, other studies suggest Also, not unexpectedly, there was a re- hood acute leukemia (39). that HDACi’s hyperacetylate NFκB. An- duction in cellular infiltration of islets other possible mechanism is the ability of with a return of normoglycemia. Al- MOLECULAR MECHANISMS OF ACTION HDACi’s to acetylate STAT-3. SAHA in- though these studies demonstrate a ben- OF HDACi’s RELATED TO creases histone H3 acetylation in the nor- eficial effect of TSA, it was reported that ANTIINFLAMMATORY PROPERTIES mal mouse brain accompanied by in- splenic T lymphocytes derived from pro- Each of the 14 reports in this issue pro- creased expression of the neuroprotective tected mice resulted in enhanced expres- poses a molecular mechanism for the proteins Hsp70 and Bcl-2. sion of IFNγ mRNA and protein (36), and ability of HDACi’s to reduce the severity Because HDACi’s increase gene expres- there were no changes in the expression of disease in animal models or in cul- sion, one possible mechanism of action is of inducible NO, IL-17 or TNFα. These tured cells. There is no single unifying the induction of genes that are themselves findings are highly inconsistent with the molecular mechanism. As reviewed by inhibitors of inflammation, for example, known beneficial effects of suppression Christensen et al. (34), NFκB received a induction of transforming growth factor of IFNγ activities, not only for the NOD great deal of attention for its role in cy- (TGF)-β, IL-10 or IL-1 receptor antagonist. mouse (37) but also for the property of tokine induction of cytokines. If one ex- Suppressors of cytokine signaling-1 and - HDACi’s to reduce IFNγ production amines the effects of IL-1β on the insulin- 3 (SOCS1/3) have broad antiinflamma- (6,10,11) and also on the property of producing pancreatic β cell, >2,000 genes tory and immunosuppressive properties, HDACi’s to suppress STAT-1 activities are modified (34). Since HDACi’s reduce and HDACi’s increase the expression of (16). Moreover, the combination of IL-1β the effects of NFκB activity- dependent SOCS-3 (41). Other studies show that plus IFNγ is highly toxic to the β cell and genes, for example IL-1-induced IL-6 or HDACi’s impair the binding of c-Fos and is prevented by SAHA and ITF2357 inducible NO synthetase (35), it seems c-Jun to DNA, thus inhibiting gene ex- (6,32). logical that the primary mechanism of pression induced by activator protein-1 action is inhibition of NFκB functions. (AP-1). EFFECT OF HDACi’s ON CHEMOKINE However, although several reports indi- Histone deacetylase 6 (HDAC-6) is a RECEPTORS cate that HDACi’s do reduce NFκB activ- mostly cytoplasmic class II HDAC that In several models of inflammation, in- ity, electrophoretic mobility shift assay has a unique structure with two catalytic filtration of neutrophils is characteristic did not reveal such a mechanism in domains and a domain-binding ubiquitin and the chemotactic chemokine IL-8 IL-1β-exposed cells in the presence of with high affinity. This HDAC acetylates plays an important role. However, de- SAHA or TSA (32). The failure to observe α-tubulin and heat shock protein spite the suppression of TNFα and IFNγ a reduction in NFκB activity was also re- (Hsp)-90. HDAC-6 was targeted by ho- gene expression in PBMCs stimulated ported for primary human synovial fi- mologous recombination, and HDAC-6- with LPS in the presence of SAHA (10) or broblasts stimulated with IL-1β for the in- deficient mice are viable, are fertile and ITF2357 (11), there is no reduction in IL-8 duction of IL-6 (24). IL-1β induces a show hyperacetylated tubulin in most gene expression or protein levels. Rather, biphasic phosphorylation of inhibitor tissues. Lymphoid development is nor- it appears that HDACi’s reduce protein IκBα with the second peak re- mal, but the immune response is moder- chemokine receptors. PBMCs from seven duced by HDACi’s (32). However, no ef- ately affected. Furthermore, the lack of healthy donors were incubated in the fect was observed on IκBα degradation HDAC-6 results in a small increase in presence of ITF2357 or valproic acid, and or NFκB binding to DNA. Thus, it ap- cancellous bone mineral density. Thus, after 4 and 24 hours, there was a de- pears that HDAC inhibition prevents the these data demonstrate that mice survive crease in CCR4 of 47% (P < 0.001) and cytokine-mediated effects by downregu- well without HDAC-6 and that tubulin 54% (P < 0.001), respectively, at therapeu- lating NFκB transactivating activity (40). hyperacetylation is not detrimental to tic concentrations (38). Consistent with In non–β cells, IFNγ-induced JAK activa- normal mammalian development. the reduction of surface CXCR4 expres- tion and STAT-1 activity depends on Protein lysine acetylation is a post- sion, steady-state mRNA levels of HDAC-1, -2 and -3 activity (16). translational mechanism involved in all CXCR4 after 2 and 4 hours, as measured Some studies demonstrate that aspects of cell homeostasis. Indeed, a by RT-PCR, were reduced by 65% in cul- HDACi’s prevent the phosphorylation of growing number of nonhistone proteins tures treated with 250 nmol/L ITF2357 IκB with downstream suppression of are targets of histone acetyltransferase compared with control cultures. There gene expression after TLR or cytokine (HAT) and HDAC. Acetylation affects

338 | DINARELLO ET AL. | MOL MED 17(5-6)333-352, MAY-JUNE 2011 REVIEW ARTICLE

thousands of cellular proteins, both cyto- deacetylates lysine 218 and 221 of p65 sponse. As expected, the MAPK cascade plasmic and nuclear (42). Among these and promotes its high affinity binding to is regulated at multiple levels, one of proteins, transcription factors are broadly IκBα in the nucleus. This complex is rap- which relies on the deactivation of its involved in the molecular mechanisms idly exported from the nucleus to reestab- components by specific phosphatases. influenced by HDACi’s. An example of lish the latent cytoplasmic pool ready to MAPK phosphatases (MKPs) are dual- acetylation of a transcription factor by be activated upon a new stimulation. specificity phosphatases that inactivate HDACi’s is STAT-3, as reviewed by Choi Maintenance of lysines 218 and 221 in an MAPK members by dephosphorylating and Reddy (43). Both ITF2357 and SAHA acetylated state prolongs the nuclear ac- phosphotyrosine and phosphothreonine induce the indoleamine 2,3-dioxygenase tivity of NFκB. Thus, from this study, it residues. In a recent report, Cao et al. (51) (IDO), which prevents dendritic cell mat- seems that preserving HDAC3 activity in demonstrated that MKP-1 is subjected to uration and is regulated by STAT-3 (44). the nucleus would lead to a blockade of regulation by reversible acetylation. STAT-3 suppresses transcription of IFNγ NFκB activity, in apparent contradiction Acetylated MKP-1 has enhanced binding and IL-17 and increases production of with the effect exerted by pan-HDACi’s to p38 and phosphatase activity, thereby IL-10. Individuals with a mutation in on proinflammatory response. reducing the level of active phosphory- STAT-3 suffer from hyper IgE syndrome, The contradictory effect of HDAC in- lated p38. The direct consequence of the resulting in failure to control inflamma- hibition on the activity of NFκB is also increased activity of MKP-1 is the damp- tion; disease severity in these patients is evident from the work of Kiernan et al. ening of TLR-mediated inflammatory re- consistent with studies showing that (49), who showed that, in a different cel- sponse. Inhibition of HDAC activity by STAT-3 suppresses cytokines. Activation lular system, the acetylation of p65 took TSA reduced the expression of TNFα, of STAT-3 can take place by acetylation, place on different lysine residues, IL-6, and IL-1β in LPS-stimulated and acetylated STAT-3 enhances STAT-3 namely Lys122 and Lys123. Moreover, macrophages in an MKP-1-dependent dimer and thereby activity. ITF2357 and the acetylation of p65 reduced its ability manner (51). In the same study, the au- SAHA induce acetylation of STAT-3 in to bind DNA and facilitated its removal thors also showed the inhibition of NOS2 dendritic cells at concentrations that are from DNA and consequently its IκB- in Raw cells by TSA. NOS2 is a transcrip- achievable in humans (44). mediated export from the nucleus. tional target of NFκB, which was not in- As a posttranslational mechanism, the HDAC-3 was shown to be the main iso- fluenced by TSA treatment. effects of lysine acetylations on cell func- form able to deacetylate p65, but in this The role of the cytoplasmic HDAC-6 in tions are broad. Over 3,600 lysine acetyla- case, inhibition of HDAC-3 would lead the regulation of inflammatory response tion sites were identified on 1,750 pro- to repression of NFκB activity. Inhibition was investigated on the basis of the ra- teins, and the acetylations due to SAHA of NFκB is not a necessary property for tionale that its inhibition would lead to and MS-275 were studied (45). It was re- an HDACi to be antiinflammatory. In an altered microtubule network and re- ported that lysine acetylation preferen- murine dendritic cells, TSA and vorinos- duced secretion of proinflammatory cy- tially targets large macromolecular com- tat strongly reduced the late expression tokines such as IL-1β (52). Although the plexes and that acetylation impaired of TNFα but did not affect IL-1β produc- necessity of a functional cytoskeleton phosphorylation-dependent interactions tion or mitogen-activated protein kinase and microtubule network was estab- (45). The reversible acetylation of key ly- activation (50). lished in this work, no clear relationship sine residues plays a fundamental role in Inhibition of proinflammatory re- between HDAC-6 inhibition and IL-1β the function of NFκB (46). The interaction sponses induced by Toll-like receptor ac- secretion was found. The effect of HDAC of NFκB (47,48) with acetyl transferase tivation by HDACi’s certainly relies on inhibition on IL-1β secretion is relevant and deacetylase enzymes has many func- multiple targets, but strongly depends on in that activation of caspase-1 via the in- tional consequences. Reversible acetyla- cell type and stimulation. The status of flammasome is required for the matura- tion of distinct lysine residues has both the same gene may differ from one cell tion of IL-1β (53). Thus, inhibition of positive and negative regulatory effects to another, and its activation or inhibi- gene expression of a critical component on nuclear NFκB activities, including tion may take place with different kinet- of the inflammasome could be envisaged transcriptional activation, DNA binding ics. The TLR4 signaling complex acti- as a mechanism responsible for the activ- affinity, IκB association and subcellular vates three main intracellular pathways: ity of HDACi’s on IL-1β secretion. More- localization. Since NFκB plays a major the NFκB signaling pathway, interferon- over, acetylation of inflammasome com- role in orchestrating the inflammatory re- related factor (IRF) signal transduction ponents could be an important sponse, its regulation by acetylation has pathways and the mitogen-activated pro- posttranslational mechanism involved in been the subject of various studies aimed tein kinase (MAPK) cascade. Inhibition its assembly and activity. at elucidating the antiinflammatory of the latter with p38 kinase inhibitors HDAC-6 contributes to the regulation mechanism of HDACi’s. Chen et al. has been proven to strongly attenuate the of TLRs and IL-1β-induced signaling. showed that histone deacetylase 3 proinflammatory innate immune re- Once again, the activity of a deacetylase

MOL MED 17(5-6)333-352, MAY-JUNE 2011 | DINARELLO ET AL. | 339 HDAC INHIBITORS IN NONCANCER DISEASES

enzyme (HDAC-6), instead of its inhibi- analyses of macrophages and DC stimu- sult in the loss of elasticity of the lung to tion, appears to be necessary to down- lated with different TLR agonists. The maintain the alveolar architecture. Alveo- modulate the proinflammatory program study revealed a critical role for HDACs lar cell death is also believed to con- induced by TLRs and the IL-1 receptor. in the expression of host defense genes, tribute to the disease. In a large clinical The study showed that HDAC-6 and se- including pattern-recognition receptors, trial in patients with COPD, histone questosome 1 (SQSTM1) were required costimulatory molecules, kinases, tran- HDAC expression and activity are re- for MyD88 aggregation and lead to inhi- scription regulators, complement factors, duced in the lung tissue (60). Mizuno et bition of TLR ligand-induced expression cytokines, chemokines and growth fac- al. (61) explored whether HDAC inhibi- of IL-6 and NOS2 in RAW264.7 cells. tors. The data suggest that HDACi’s exert tion causes lung cell apoptosis and em- HDAC-6 and SQSTM1 partially sup- mainly an immunosuppressive effect but physema in a rat model. Rats were pressed the activation of p38 and JNK, also possess immunostimulatory activi- treated with TSA. Chronic TSA treatment but they had no effect on degradation of ties. Interestingly, they suggest a novel ac- increased the alveolar air space area and IκB (54). According to these findings, in- tivity of HDACi’s, namely the induction the number of caspase-3–positive cells in hibition of HDAC-6 would lead to en- of the chromatin modifier Mi-2beta. Mi-2 rat lungs. TSA suppressed hypoxia- hanced MyD88 signal transduction activ- is part of the nucleosome remodeling, his- inducible factor-1α and vascular en- ity. This result would be the logical tone deacetylation (NuRD) complex. The dothelial growth factor (VEGF) gene ex- consequence of the use of an HDAC-6 in- Mi-2β complex is selectively recruited pression in rat lungs as well as cultured hibitor, such as vorinostat, in a LPS- along with the other complexes to the human pulmonary microvascular en- induced septic shock model. A recent re- control regions of secondary response and dothelial cells. The study suggests that port (55), showed that administration of delayed primary response inflammatory administration of HDACi’s may worsen vorinostat (50 mg/kg intraperitoneally) genes. In these complexes, the Mi-2β the clinical course of COPD. after a lethal dose of LPS significantly complex acts antagonistically to limit the improved long-term survival and re- induction of these two gene classes. DOES THE USE OF HDACi’s INCREASE duced the expression of TNFα and IL-6. Mi-2β/NuRD inhibits LPS-induced sec- THE RISK OF INFECTIONS? The authors also observed that vorinos- ondary and delayed primary cytokine The inhibition of a number of proin- tat reduced the increased expression of production in macrophages and dendritic flammatory cytokines induced by micro- MyD88 caused by LPS administration cells (58). Thus, HDACi’s inhibit the ex- bial products through TLR’s activation and postulated that this may be the di- pression of secondary response genes by HDACi’s poses a question related to rect consequence of the decreased ex- such as IL-12, IL-6 and inducible NOS, the effect of the use of these inhibitors on pression of MyD88-associated proinflam- but have no effect or delayed effect on the occurrence of infections. Without a matory genes. primary genes such as IL-1β and TNFα, doubt, all biologics used in patients with A large number of studies have shown respectively. Roger et al. also found that various inflammatory diseases increase that inhibition of HDAC activity leads to TSA increased histone H4 acetylation of risk of infection. Anti-TNFα monoclonal repression of different TLR-induced re- IL-6 and IL-12 promoters, suggesting that antibodies (infliximab and adalimumab) sponses, mainly on the basis of analysis there was no correlation between the sta- and the soluble TNF receptor etanercept of the expression of proinflammatory cy- tus of histone acetylation and the ob- are associated with increased risk of op- tokines. Two reports have analyzed the served downregulatory effect of TSA on portunistic as well as routine infections effect of HDAC inhibition using a more gene transcription. and deaths (reviewed by Dinarello [1]). comprehensive approach on the basis of Blocking IL-1β with anakinra is not asso- whole genome expression analysis. Brog- ARE HDACi’s CONTRAINDICATED IN ciated with increased opportunistic infec- don et al. (56) showed that the potent SOME DISEASES? tions and is considerably safer. Some bio- pan-HDACi LAQ824 exhibited a rather The third cause of death after cardio- logics are linked to progressive selective activity on macrophages and vascular disease and cancer is chronic multifocal leukencephalopathy. At the dendritic cells stimulated with LPS. The obstructive pulmonary disease (COPD). present time, there are only two disease inhibitor specifically inhibited DC-con- Patients with COPD have a progressive, conditions that are being treated with trolled T helper 1 (Th1) effector but not unrelenting course, and various thera- HDACi’s: vorinostat in graft versus host Th2 effector cell activation and migra- pies have not significantly reduced dis- disease (GvHD) (43) and givinostat in tion. It also inhibited macrophage- and ease severity. Cytokines such as TNFα, systemic-onset juvenile idiopathic arthri- DC-mediated monocyte but not neu- IL-18 and IL-32 are thought to play a role tis (SOJIA) (5). The GvHD study is un- trophil chemotaxis. in that these cytokines are elevated in derway and there are no data at this Roger et al. (57) confirm the concept of lung tissue of smokers with COPD but time, but there were no infections in the HDACi’s as antiinflammatory drugs not in smokers without COPD (59). SOJIA patients treated for 12 weeks with through use of genome-wide microarray Chronic inflammation is believed to re- givinostat. Moreover, these patients with

340 | DINARELLO ET AL. | MOL MED 17(5-6)333-352, MAY-JUNE 2011 REVIEW ARTICLE

SOJIA were concomitantly treated with with a history of dyslipidemia (P = 0.05) levels of ITF2357 were in the range of glucocorticoids at the same time. In the (68). The data from humans as well as 125–200 nmol/L. There was no organ phase I trial of givinostat, the decrease in mice support the hypothesis from the toxicity, and the study achieved signifi- whole blood production of TNFα, IL-1β, large epidemiologic studies that IL-1β is cant (P < 0.01) reduction in parameters of IL-6 and IFNγ was transient after a single mainly involved in the functional alter- systemic disease as well as the number oral dose and returned to baseline levels ations of cardiomyocytes (69). It is un- of painful joints in over 75% of patients. within 12 hours. In contrast, antibodies likely that HDACi’s will be used as a At week 4, the American College of to TNFα persist for several weeks and monotherapy in combating heart failure. Rheumatology (ACR Pediatric) 30, 50 or are more likely to suppress host defense As outlined by McKinsey (63), there are 70 improvement was 77.8%, 55.6% and than short-acting oral HDACi’s (62). several drugs used to treat heart failure 22.2% and increased further to 77.8%, but few target inflammatory pathways. It 77.8% and 66.7%, respectively, at week HDACi’s FOR HEART FAILURE is also likely that HDACi’s may have an 12. The most consistent finding was the As reviewed in the report by McKin- effect on cardiac remodeling indepen- reduction in the number of active joints sey (63), there are several animal models dent of inflammatory cytokines. and/or joints with limited range of mo- that result in the physiological character- tion (5). Also, in that trial, there was a re- istics of human heart failure. Some are JOINT DISEASES duction in the white blood cell counts, related to inflammation associated with In addition to cytokine-driven loss of the percentage of neutrophils, the levels cytokine-induced cell death and the loss the insulin-producing β cell, the chon- of IL-1α and CD40 ligand as well as sedi- of myocytes, particularly secondary to drocyte is important in the production of mentation rate and CRP (26). After 12 IL-1β (64,65). As a result, the ability of proteoglycans for maintenance of carti- weeks, givinostat resulted in significant the heart to effectively pump blood out lage integrity, but is a target for cytokine- safety and benefits. of its chambers is reduced, and the pa- driven destruction (70). Overexpression tients experience difficulties maintaining of IL-1β results in irreversible cartilage GRAFT VERSUS HOST DISEASE normal activities. Heart failure is an destruction. ITF2357 is protective on car- After allogeneic bone marrow trans- enormous economic burden on most so- tilage catabolism in a mouse model of plantation in mice, treatment with cieties, and with increasing numbers of nonautoimmune inflammatory arthritis HDACi’s provides a survival benefit in patients with Type 2 diabetes and in- induced by Streptococcus pyogenes cell models of GvHD. Administration of creased myocardial infarction, the inci- walls. Administered orally at a dose of 1 SAHA after bone marrow transplanta- dence of heart failure increases each year. and 10 mg/kg, ITF2357 suppressed joint tion reduced intestinal histopathology, Having orally active HDACi’s as antiin- swelling (25). As reviewed by Joosten et clinical severity and mortality from acute flammatory agents is an attractive thera- al. (25), there was an inhibition of chon- GvHD (71). Moreover, in mice receiving peutic option. The concept that TNFα drocyte proteoglycan synthesis of 44% lethal doses of tumor cells, administra- contributed to heart failure was based on with 1 and 10 mg/kg ITF2357. Thus, tion of SAHA did not impair graft versus two observations: levels of TNFα were HDACi’s may be used to preserve carti- tumor activity and resulted in signifi- elevated in the circulation of patients lage to protect against osteoarthritis. It is cantly improved tumor-free survival. with stage III and IV New York Heart likely that modulation of IL-1β by Similar observations were reported for Association heart failure classification HDACi’s, as shown in several reports, other HDACi’s such as ITF2357 (44). The and treating heart failure patients with results in reduced inhibition of chondro- mechanism for these effects in GvHD etanercept (a TNFα blocker) improved cyte proteoglycan synthesis and hence in may be due to decreased cytokine pro- exercise tolerance (66). However, in two a protective effect against cartilage catab- duction, as was shown in those studies large studies, etanercept failed to provide olism during arthritis. As reviewed by (71). An early step in models of GvHD is improvement and infliximab actually in- Joosten et al. (25), ITF2357 also protects antigen maturation of dendritic cells. creased mortality (67). Therefore, when bone loss induced by IL-1β. Dendritic cells exposed to either SAHA there is use of HDACi’s to combat heart The proof of principle study to test the or ITF2357 increase the expression of failure, a reduction in TNFα levels is not hypothesis that HDACi’s would benefit IDO, a suppressor of dendritic cell func- considered beneficial. joint disease can be found in the observa- tions (44). As reviewed by Choi and On the other hand, studies have exam- tions of the trial of givinostat (ITF2357) Reddy (43) in this issue, acetylation of ined a role for IL-1β in the pathogenesis in children with SOJIA (5). ITF2357 is a STAT-3 activates its functions, and in- of heart failure. Serum levels of IL-1β in hydroxamic acid–containing, orally ac- deed HDACi’s activate STAT-3 by acety- a large and representative population tive inhibitor targeting class I and II lation (72). With the acetylation of STAT- were associated with congestive heart HDAC. The children received a daily 3, there is downregulation of dendritic failure (P > 0.001) and angina (P = 0.02), dose of 1.5 mg/kg in two divided doses. cells. Therefore, HDACi’s likely serve with Ca2+ serum levels (P = 0.02) and At a dose of 1.5 mg/kg, the peak blood several functions in the prevention of

MOL MED 17(5-6)333-352, MAY-JUNE 2011 | DINARELLO ET AL. | 341 HDAC INHIBITORS IN NONCANCER DISEASES

GvHD. STAT-3 also promotes the tran- duce IL-1β-driven death of the insulin- for HDAC-1 and HDAC-2, revealed no scription of IDO. Thus, acetylation of producing β cell in primary rat and reductions in atherosclerotic lesions (80). nonnuclear proteins such as STATs pro- mouse islets and cell lines (6,32,35) as re- In fact, treatment with TSA worsened the vides mechanistic insight into HDAC in- viewed in this issue of Molecular Medicine disease. TSA was injected intraperi- hibition of cytokine production and im- (34). Similar studies were reported for in- toneally every other day at 0.5 and mune cell functions (72). In terms of hibition disease in animal models of 1.0 mg/kg and, as stated by the authors, cellular mechanisms, inhibitors of HDAC rheumatoid arthritis and osteoarthritis, as at “doses used to treat tumors in mice” increase Fox-p3–expressing T-regulatory reviewed in this issue (25). At first glance, (80). The 50% inhibitory concentration of cells, which contribute to the reduction these are seemingly unrelated models. But TSA for HDAC-1 is 2 nmol/L and for of disease in autoimmune models. SAHA on closer examination, each has inflam- HDAC-2 is 3 nmol/L (7). Thus, these is presently in trials to prevent or reduce matory cytokines as part of the pathologi- doses are likely to be considerably higher disease severity in GvHD in patients cal mechanism. For example, there is a cy- than those required to suppress inflam- with bone marrow transplants. The topic tokine pathway in established models of mation. In vitro, acetylation of H4 in of HDACi’s in GvHD is reviewed in de- inflammatory bowel disease (74), lupus mouse macrophages was observed at tail in this issue (43). nephritis (75), chronic heart failure (63), 10 ng/mL (33 nmol/L), which is 10-fold

rheumatoid arthritis and osteoarthritis greater than the IC50 for HDAC-2 (80). WHERE DO HDACi’s FIT INTO THE (25), multiple sclerosis (76) and GvHD Cyclooxygenase-2 (COX-2) is consid- CURRENT TREATMENT OF CHRONIC (43). There is also a cytokine component ered proatherogenic, and COX-2 expres- INFLAMMATORY DISEASES? in traumatic brain injury (77). However, it sion is in increased bone marrow– is unlikely that inflammation affects dis- derived mouse macrophages treated Joint, Bowel and Metabolic Diseases ease in muscular dystrophy (78). with SAHA or TSA. Similarly, plasmino- As stated above, the antiinflammatory gen activator inhibitor-1 (PAI-1) is also properties of HDACi’s are based prima- Atherosclerosis proatherogenic, and its expression is in- rily on a reduction in either the produc- More than any other chronic inflamma- creased in the same macrophages treated tion or the activity of inflammatory cy- tory disease, atherosclerosis is the major with SAHA or TSA (81). However, the tokines in animals as well as in primary cause of death in the developed world concentration of SAHA added to these cells at significantly lower concentrations because of overnutrition and Type 2 dia- cultures was 10 μmol/L, which is 100- that those used in tumor cells. If betes; overnutrition is also a cause of fold higher than the IC50 for HDAC-1 HDACi’s are to be used for treating Type 2 diabetes in developing countries. and 75-fold higher than that for HDAC-2 chronic inflammatory conditions, Because of their safety and ease of oral (7). In mouse macrophages stimulated HDACi’s should also inhibit the produc- administration, most patients at high risk with LPS, PAI-1 protein levels increase tion or activities of mediators such as ma- of a cardiovascular event are treated with over 10-fold when incubated with TSA; trix metalloproteinases, cyclooxygenases, a “statin,” since these drugs have re- however, the concentration of TSA used adhesion molecules and chemokines. As duced the incidence of cardiovascular in those studies was 500 nmol/L, which of this writing, the only clinical trial to events and deaths. The beneficial view of is a 250-fold higher concentration than support the concept that a low dose of statins is thought to be because of a com- the IC50 for HDAC-1 (81). Similarly, HDACi’s exhibits antiinflammatory prop- bination of their property to reduce low- COX-2 increased over 40-fold in the pres- erties is the trial of ITF2357 in children density lipoproteins (LDLs) as well as an- ence of 500 nmol/L TSA, which is 250- with SOJIA (5). ITF2357 (generic givinos- tiinflammatory agents. By comparison, fold over the IC50 for TSA on HDAC-1. tat) is a hydroxamic acid–containing, HDACi’s could be ideal for reducing in- Importantly, at 10 nmol/L TSA, which is orally active inhibitor targeting zinc-de- flammation in the atherosclerotic plaque. only three-fold greater than the IC50, pendent HDAC isoform class I and II As pointed out in the review by Halili et there was no increase in COX-2 expres- HDAC. On the basis of a reduction in the al. (79), little is known about the effects of sion in LPS-stimulated macrophages (81). secretion of IL-1β, givinostat was also HDACi’s on the atherosclerotic process. In addition, an increase in PGD2 by used to treat Schnitzler syndrome, as re- There are two mouse models of athero- HDACi inhibits the production of IL-1- ported in this issue of Molecular Medicine sclerosis: the apoenzyme E (APOE) and induced metalloproteinases (82). In con- (73). Vorinostat is presently in clinical tri- the LDL receptor–deficient mouse. To our trast, Yamaguchi et al. (83) reported that als in patients with GvHD (43), on the knowledge, there is one study in which HDACi’s block the induction of c-jun basis of animal models (44,71,72). HDACi’s was used to treat these mice transcription by inhibiting the recruit- In addition to inhibition of the produc- and evaluate the effects. ment of the complex to the c-jun pro- tion of inflammatory cytokines, HDACi’s LDL-deficient mice fed a high-fat diet moter. This results in reduced expression also inhibit cytokine post-receptor signal- for 4 and 8 weeks and treated with TSA, of several activator protein-1– dependent ing. For example, SAHA and ITF2357 re- a specific histone deacetylase inhibitor genes, including COX-2.

342 | DINARELLO ET AL. | MOL MED 17(5-6)333-352, MAY-JUNE 2011 REVIEW ARTICLE

Similar data can be found in an earlier never be known. It is interesting to note patients with SOJIA, clinical and hemato- report. Bone marrow–derived mouse that the beneficial effect of TSA in the logical improvement is observed at this macrophages were stimulated with LPS NOD model was achieved at a dose of low dose (5,26) and similarly in patients and at various time points, HDAC-1, 0.5 mg/kg once weekly. However, the with Schnitzler syndrome (73). It ap- HDAC-4, HDAC-5, HDAC-7, HDAC-8 reduction in diabetes was only observed pears that HDACi’s such as ITF2357 and HDAC-9 were elevated (84). Of when the treatment was limited to the (givinostat) may be candidates for treat- these, HDAC-1 exhibited the greatest in- interval of 18–24 weeks of age (36), ing chronic inflammatory diseases in- crease and duration. TSA at a concentra- whereas early treatment at 5–18 weeks cluding atherosclerosis. tion of 500 nmol/L was examined for its of age was without effect. In the LDL effects on LPS-induced genes. With use receptor–deficient mouse, for inducing TREATING ACUTE DISEASES WITH of microarray analysis of LPS-inducible atherosclerosis, the dose was 1.0 mg/kg HDACi’s genes, one group was unaffected, an- every other day. One may conclude that Studies have focused on mouse mod- other was superinduced and a third the antiinflammatory properties of els of sepsis, but it is unlikely that group was suppressed. The genes that HDACi’s are best observed at low dos- HDACi’s will be tested in clinical trials were suppressed were not revealed in ing schedules and that few conclusions for human sepsis. HDACi’s have also the report. Of those genes that were su- can be drawn by using antitumor doses been tested in models of ischemia/reper- perinduced, COX-2 mRNA and protein in mice. However, one may also inter- fusion (88,89). Ischemia/reperfusion is increased. A similar increase was ob- pret from these data that treating ro- essentially sterile inflammation, and, served for the chemokine CXCL2 and dents with HDACi’s in models of dis- here, cytokines such as IL-1α and TNFα endothelin (84). The authors concluded ease requires higher doses than in play a significant role. But the testing of that HDACs act as negative regulators of humans. HDACi’s in humans with hemorrhagic LPS-induced COX-2 (84). In support of Unlike TSA, there are ample data in shock, acute stroke or myocardial infarc- this concept, mouse macrophages were studies with SAHA, FK228 or ITF2357 in tion is doubtful. One recurrent acute dis- transfected with the proximal promoter humans. Blood levels and tolerability ease is gout and, here, HDACi’s may of COX-2, and LPS increased the activity have been established. In the case of provide an additional benefit to of the promoter five-fold. In the pres- SAHA, the maximal tolerated dose of colchicine, as studies demonstrate that ence of TSA (500 nmol/L), the increase 800 mg/day in two divided doses in- gout is highly responsive to IL-1β block- was 18-fold (84). duces fatigue, nausea and some vomiting ade (90–92). The findings of increased IFNγ and (85,86). In the report by Halili et al. (81), TNFα production in the study in NOD SAHA increased gene expression of NEURODEGENERATIVE DISEASES mice (36), the exacerbation of atheroscle- COX-2, PAI-1 and IL-12 in mouse Neurodegeneration includes the slow rosis in the LDL-deficient mouse (80), macrophages at 10 μmol/L, a concentra- progressive forms characteristic of ather- the increases in VCAM-1 and scavenger tion 100-fold greater than the IC50 for osclerotic brain disease as well as Alz- receptor CD36 (80) in COX-2 (81, 84) and HDAC-1 (7), which is clinically irrele- heimer disease but also that following elevated PAI-1 (81) by TSA and SAHA in vant. In vitro, SAHA reduces LPS- acute brain trauma of a stroke or blunt mouse macrophages are inconsistent induced cytokines in primary human injury. Histone deacetylation is often al- with the overall portfolio of antiinflam- PBMCs at concentrations between 50 and tered in the brains of animals used to matory properties of HDACi’s. How 200 nmol/L, far lower than the 10 μmol/L study neurodegeneration. In this issue, does one explain the discrepancies? used by Halili et al. The concentration of Shein and Shohami (77) review the evi- Using in vitro concentrations of TSA of 10 μmol/L in vitro is used to induce dence for the use of HDACi’s in treating 500 nmol/L is certainly inconsistent proapoptotic genes in tumors. For either the acute or slowly progressive with the low concentration of this ITF2357, there is tolerability in healthy course of neurodegeneration. The effi-

HDACi with an IC50 for HDAC-1 and humans (87), in patients with autoin- cacy of HDACi’s in acute brain injury HDAC-2 of 2 and 3 nmol/L, respectively flammatory diseases (73) and in children was reported by Shein et al. (93) as well (7). LDL receptor–deficient mice were with SOJIA (5). At an oral dose of as in models of amyotrophic lateral scle- treated with 1 or 0.5 mg/kg of TSA 1.5 mg/kg, ITF2357 (givinostat) is well rosis (ALS), Alzheimer disease, spinal every other day, the dose used to treat tolerated. At this dose, there is a transient muscular atrophy and experimental au- tumors in mice. In the case of studies reduction in LPS-induced cytokines in toimmune encephalomyelitis. Although with TSA, one cannot predict the rele- whole blood cultures during a phase I trial many neurodegenerative diseases in- vance for human disease, since there are in healthy humans, and this reduction oc- clude a cytokine-driven inflammatory no studies on the blood levels of TSA curs at a blood level of 125–250 nmol/L, component, Huntington disease remains after intraperitoneal injections, and the concentrations known to inhibit LPS- a uniquely noninflammatory loss of neu- toxicity of TSA in humans will likely induced cytokines in PBMCs (11,87). In rons. Nevertheless, several models of

MOL MED 17(5-6)333-352, MAY-JUNE 2011 | DINARELLO ET AL. | 343 HDAC INHIBITORS IN NONCANCER DISEASES

neurodegeneration are due to inflamma- crosses the blood-brain barrier and in- genic mouse expressing mutant superox- tory processes. In this regard, the antiin- creases histone acetylation in the brain ide dismutase 1 (SOD1). Mutant SOD1 flammatory properties of HDACi’s ap- (96). The administration of SAHA has activates caspase-1 and IL-1β in microglia pear to be neuroprotective. In animal consistently shown this therapeutic ben- (100) and supports the concept that neu- models of Alzheimer disease, the most efit in the mouse model; nevertheless, roinflammation contributes to ALS dis- studied neurodegenerative disease, which HDAC is affected by SAHA to ease progression (101). Mutant SOD1- HDAC inhibition decreased β-amyloid improve motor function in the R6/2 induced IL-1β correlated with amyloid-like production and effective neuroprotection mouse model remains unclear. Brain misfolding and was independent of dis- was associated with increased acetyla- levels of HDAC-7 are reduced in both mutase activity. ALS-linked cytoplasmic tion of histone 4 (94). In a model of wild-type animals and in animals with accumulation of mutant SOD1 was sensed closed head trauma, the antiinflamma- SAHA (97). However, there was no im- by an apoptosis-associated speck-like tory properties of HDACi’s prevented provement in a number of physiological protein containing a C-terminal caspase the deterioration of neurological impair- or behavioral phenotypes in R6/2 mice recruitment domain (ASC)-containing ment, even when administered 24 hours deficient in HDAC-7 (97). Thus, the inflammasome and antagonized by after the injury (93). SAHA-mediated amelioration or disease autophagy, limiting caspase-1–mediated severity in the mouse model for Hunt- inflammation (100). A deficiency in Huntington Disease ington disease is not via HDAC-7, and caspase-1 reduced IL-1β levels, extended Although the role of inflammation in the ability of SAHA to reduce brain lev- the lifespan of G93A-SOD1 transgenic Huntington disease is presently unclear, els of HDAC-7 appears to be unrelated. mice and attenuated inflammatory inhibition of caspase-1 delays the onset Intracerebroventricular administration pathology. Similar results were observed of motor abnormalities in a mouse model of a caspase inhibitor delays disease pro- with treatment with recombinant IL-1Ra of the disease (95) and treatment with gression and mortality in the transgenic (100). These findings identify microglial oral SAHA also reduced disease severity mouse (95). In a rat cerebral artery occlu- IL-1β as a causative event of neuroinflam- (96). Because the effect of valproic acid sion model of stroke, post-insult treat- mation and suggest IL-1 as a potential on the central nervous system is well ment with the HDACi valproic acid re- therapeutic target in ALS (101). documented, clinical trials with oral duced ischemia-induced brain infarction, HDACi’s could be initiated in patients caspase-3 activation and neurological IS THERE A ROLE FOR CYTOKINE with early indications of dementia. Al- deficits (98). In a rat excitotoxic model of INHIBITION IN MULTIPLE MYELOMA? though currently available HDACi’s are Huntington disease, treatment protected The cytokine IL-6 is the primary growth “pan HDAC” inhibitors, HDACi’s with against DNA damage, caspase activation factor for myeloma, and anti–IL-6 mono- greater specificity will be the future di- and apoptosis of striatal neurons (98). clonal antibodies as well as anti–IL-6 re- rection for this class of drugs. Thus, HDACi’s are potential drugs for ceptor monoclonal antibodies are used to Huntington disease is an autosomal- treating some forms of neurodegenera- treat Castleman disease. However, in pa- dominant genetic disease. The patholog- tive diseases, and whether the benefit of tients with a premyeloma condition called ical lesions occur in the striatum and HDACi’s is via a cytokine-mediated pro- smoldering or indolent myeloma, block- cortex and the disease is fatal approxi- cess such as caspase-1 remains to be ing IL-1β to reduce IL-6 production signif- mately 20 years after the onset of symp- demonstrated. In neurons, acetylation at icantly prevents progression to full-blown toms. There is an expanded CAG/ lysine 40 of α-tubulin increased by TSA multiple myeloma (102,103). HDACi’s in- polyglutamine repeat in “huntingtin,” a results in the release of the brain- derived hibit the production of IL-6 by LPS as protein of unknown function. The dis- neurotrophic factor (99). Tubulin acetyla- well as IL-1β-driven IL-6 (10,11,23,24). ease is progressive with loss of motor tion is reduced in the brains of animals Therefore, although multiple myeloma is function, and there is no treatment to with Huntington disease, and TSA com- cancer, the use of HDACi’s in the precan- slow the progress of the disease. In pensates for the transport- and release- cer condition of smoldering/indolent my- 2003, there was a report that SAHA defect phenotypes that are observed in eloma is a likely therapy compared with added to the drinking water of R6/2 disease (99). the daily injection of anakinra that was HD mice, the model for this disease, re- used in the trial (102). sulted in marked improvement in motor Amyotrophic Lateral Sclerosis The activity of ITF2357 was studied on function (96). Transgenic mice express- ALS is another disease in which there multiple myeloma and acute myeloge- ing human huntingtin with an ex- is progressive loss of brain neurons. A nous leukemia cells in vitro and in vivo. panded CAG/ polyglutamine repeat de- study in mice with mutations in super - ITF2357 induced apoptosis in freshly ob- velop a progressive syndrome with oxide dismutase revealed a role for tained multiple myeloma and acute β many of the characteristics of human caspase-1 processing of IL-1 . The mouse myelogenous leukemia cells at an IC50 of Huntington disease (95). Indeed, SAHA model for ALS is G93A-SOD1, a trans- 200 nmol/L (104). ITF2357 induced hy-

344 | DINARELLO ET AL. | MOL MED 17(5-6)333-352, MAY-JUNE 2011 REVIEW ARTICLE

peracetylation of histone H3, H4 and in both cell lines and freshly isolated pa- ers of systemic inflammation such as re- tubulin. ITF2357 was also cytotoxic for tient cells (108). The decrease in the IL-6 current fevers. The low concentrations in the IL-6–dependent multiple myeloma receptor expression was accompanied by vitro and in vivo also support the concept cell line CMA-03. Of particular relevance, decreased signaling, as measured by that the cytokine-reducing properties of ITF2357 inhibited the production of IL-6, STAT-3 phosphorylation. These data sup- HDACi’s contribute to controlling dis- vascular endothelial growth factor and port the previous studies that ITF2357 in- ease activity rather than direct tumor cell IFNγ by mesenchymal stromal cells by hibits cytokine-mediated myeloma cell death. To accomplish tumor cell death, 80–95% (104). In vivo in mice, ITF2357 sig- growth and survival. dexamethasone or another proapoptotic nificantly prolonged survival of severe regimen is required. combined immunodeficient mice inocu- IS THERE A ROLE FOR CYTOKINE lated with acute myelogenous leukemia INHIBITION IN MYELOPROLIFERATIVE CUTANEOUS T-CELL LYMPHOMA cells at a dose of 10 mg/kg and inhibited DISORDERS? Two HDACi’s, vorinostat and ro- the production of growth and angiogenic The efficacy of givinostat in 12 patients midepsin, are approved for the treatment factors by bone marrow stromal cells, in with polycythemia vera and in 16 pa- of cutaneous T-cell lymphoma (CTCL). It particular IL-6 and VEGF (104). These re- tients with myelofibrosis, bearing the is likely that there is a role for cytokines sults suggest that part of the mechanism JAK2V617F mutation, has been reported in this disease. The levels of IL-1α in epi- of action of HDACi’s in multiple mye- (109). Givinostat was given orally for 24 dermis of patients with CTCL was inves- loma and acute myelogenous leukemia weeks at a dose of 50 mg twice daily; tigated by using immunohistochemistry may be due to a reduction in cytokine however, a dose reduction was applied and specific enzyme-link immunosorbent production, since these hematopoietic in 10 patients. The median treatment du- assays. Significant increases of IL-1α malignancies are driven by IL-1β and ration was 20 weeks. The troubling puri- were released from the involved CTCL IL-6 as growth factors (102,105,106). A tus disappeared in most patients, and re- epidermis compared with the normal phase II, multiple-dose clinical trial in 19 duction of splenomegaly was observed epidermis from healthy individuals (111). patients with relapsed or progressive in 75% of polycythemia vera patients Both keratinocytes and leukocytes could multiple myeloma was carried out (107). and 38% of myelofibrosis patients. Re- release IL-1α, but the majority was de- The first six patients received 150 mg verse transcription polymerase chain re- rived from the keratinocytes. The IL-1α ITF2357 twice daily, but because of toxic- action identified a trend to reduction of was biologically active (111). Extracorpo- ity, the dose was reduced to 100 mg the JAK2V617F allele burden. Givinostat real photopheresis is used to treat CTLC. ITF2357 twice daily. This was the maxi- was well tolerated and could induce Following this therapy, lymphocytes be- mal tolerated dose. Although these pa- hematological response in most poly- come apoptotic and monocytes, exposed tients had advanced disease, a dose of cythemia vera patients (109). In vitro, the to post-extracorporeal photopheresis 100 mg twice daily alone or combined effect of ITF2357 was evaluated on carry- (post-ECP), reduce spontaneous proin- with dexamethasone, ITF2357 proved tol- ing the JAK2 (V617F) mutation obtained flammatory cytokine secretion including erable and showed a modest clinical ben- from polycythemia vera and essential IL-1α (112). IFNγ is commonly used to efit (107). It appears that the optimal use thrombocythemia patients (110). The treat CTLC. The mechanism of action of HDACi’s in multiple myeloma would clonogenic activity of JAK2 (V617F) mu- of IFNγ in CTLC may include the induc- be in smoldering myeloma, where the tated cells was reduced by low concen- tion of IL-1Ra (113). SAHA as well as combination of IL-1β blockade plus dex- trations of 1–10 nmol/L of ITF2357, givinostat do not suppress IL-1Ra (11) amethasone significantly delays or even which is 100- to 250-fold lower than con- and, in some studies, increase the pro- prevents the progression to full-blown centrations that are traditionally required duction of IL-1Ra (87). multiple myeloma (102). to inhibit the growth of tumor cells lack- Gene expression profiling of the ing this mutation. By Western blotting, SAFETY AND TOLERABILITY KMS18 multiple myeloma cell line was ITF2357 resulted in the disappearance of REQUIREMENTS OF HDACi’s FOR performed with ITF2357 (108). The mod- total and phosphorylated JAK2 (V617F) TREATING CHRONIC DISEASES ulation of several genes and their biolog- as well as pSTAT5 and pSTAT3 (110). When considering the property of ical consequence were verified in a panel Together, these studies provide evi- HDACi’s in the treatment of chronic dis- of multiple myeloma cell lines and cells dence that, in some hematopoietic disor- eases characterized by inflammatory, au- freshly isolated from patients by using ders, the efficacy of HDACi’s is in part toimmune or degenerative processes, use polymerase chain reaction analysis and due to a reduction in either cytokine pro- of HDACi’s to treat such diseases is safe Western blotting. Out of 38,500 human duction or cytokine signaling. Clinically, and well tolerated. In general, HDACi’s genes, several were downregulated and, acute myelogenous leukemia, poly- are safe in humans. For example, val- in particular, ITF2357 down-modulated cythemia vera and multiple myeloma are proic acid, a carboxylate HDACi, is the the IL-6 receptor α transcript and protein associated with elevated CRP and mark- drug of choice for chronic therapy of

MOL MED 17(5-6)333-352, MAY-JUNE 2011 | DINARELLO ET AL. | 345 HDAC INHIBITORS IN NONCANCER DISEASES

generalized and focal epilepsy as well as as glucocorticoids. In the children with MUSCULAR DYSTROPHY obsessive disorders (114,115). Valproic SOJIA, givinostat was used in combina- As reviewed in this issue of Molecular acid has also been used to increase latent tion with steroids. Medicine (78), HDACs regulate muscle- HIV-1 gene expression in patients while In assessing the effects of various specific gene transcription via control of being treated with anti–HIV-1 drugs HDACi’s in animal models of disease, histone acetylation. Thus, the balance without interference (116). Butyrates are one must establish that any beneficial ef- between acetylation and deacetylation is used to treat patients with sickle cell ane- fect in the model is consistent with a a critical determinant of muscle gene mia and β-thalassemia (117). In fact, val- dose that is likely to be tolerated in hu- transcription and is emerging as a possi- proic acid is used chronically by millions mans. In many models, TSA is used to ble intervention for the regenerative po- of people. SAHA (vorinostat) is presently assess a role for HDACi’s in the amelio- tential of muscle stem cells. For muscu- being evaluated for treatment in GvHD ration of disease activity. For example, lar dystrophy, Consalvi et al. (78) (71). ITF2357 (givinostat) has been used TSA in models of lupus nephritis TSA showed that the mechanism responsible in children with SOJIA for 12 weeks was administered for 5 weeks at a dose for the activity of HDACi’s in the animal without adverse effects (5,26). Givinostat of 0.5 mg/kg/day with a clear beneficial model of muscular dystrophy goes be- has also been tested in patients with the effect (119). However, it is impossible to yond follistatin induction. There is the autoinflammatory disease Schnitzler syn- draw any conclusions as to whether the known potential of dystrophic muscles drome, and the observations are reported tissue concentrations of TSA required to activate mechanisms of compensatory in this issue (118). In both the SOJIA and for the effect of TSA are in the range of regeneration. This evidence supports the Schnitzler syndrome studies, there were clinically relevant HDACi’s. In this current view that HDACi’s are emerging clear indications of the desired antiin- model of lupus nephritis, mice were also candidate drugs for pharmacological in- flammatory effects in that clinical scores treated with SAHA at a dose of 25 and terventions in muscular dystrophies and of disease severity as well as a decrease 50 mg/kg intraperitoneally for 10 weeks reveals unexpected common beneficial in the circulating number of white blood (120). In humans with leukemia, the outcomes of pharmacological treatment. cells occurred. In both studies, the dose maximum tolerated dose for SAHA was The activation of a differentiating pro- of 1.5 mg/kg was well tolerated. 200 mg twice daily (1.5 mg/kg twice a gram in the stem cells resident in the Chronic diseases that are likely to be day) (85). Dose- limiting toxicities were muscle is a further key mechanism that treated with HDACi’s are rheumatoid fatigue, nausea, vomiting and diarrhea. contributes to the activity of these com- arthritis, osteoarthritis, inflammatory Even assuming that the mouse requires a pounds. Differentiation of stem cells to- bowel disease and Type 2 diabetes and, 10-fold greater amount of SAHA than ward a specific lineage through epige- being orally effective, HDACi’s have humans to reach the effective concentra- netic modulation has a potential in considerable advantages over the par- tion in vivo, SAHA at doses of 2.5 mg/kg many field of medicine. With safety on enterally administered anticytokines. was not well tolerated. its side, it is possible that a clinical trial These diseases are dramatically affected On the other hand, the use of HDACi’s of HDACi’s may result in activating by anticytokine antibodies and soluble gained considerable validity when the ef- stem cells to differentiate into muscle receptors, which have no organ toxicities fective concentration in vitro was attain- cells, but the specificity of HDAC and other than the pain of the injection able in humans after oral administration. cofactors for differentiation are not needle; indeed, they are remarkably well In the phase I trial of givinostat in healthy presently established. tolerated. In contrast, HDACi’s do have male subjects, a single oral dose of organ toxicities and, at doses used to 100 mg resulted in a peak plasma level AUTOPHAGY AND HDACi’s treat cancer, all HDACi’s cause gastroin- of 207 nmol/L at 2 hours and an area testinal disturbances and other constitu- under the curve (AUC) of 1,410 nmol/L/h Autophagy tional symptoms. These constitutional with a half-life of 6 hours (87). In vitro, In general, autophagy is an adaptive symptoms would not be tolerated in pa- 100 nmol/L of ITF2357 (givinostat) sup- mechanism protecting the loss of cells by tients without life-threatening cancer. If pressed by 90% IL-1β-driven loss of bone reusing intracellular components that clinical trials of HDACi’s are proven to in the rat calvarium model for osteoclast have been generated by pathological reduce activity in chronic diseases, they activation and similarly for TNFα-driven processes. Autophagy is often studied in will need to be well tolerated. Therefore, bone loss (25). The oral dose of 100 mg vitro in cells deprived of nutrients, but in the dose required to reach an effective also suppressed LPS-induced TNFα, IL-1β vivo, autophagy replenishes cells after an tissue level should be without significant and IFNγ 4 hours after dosing. In freshly ischemic or stressed event and prolongs constitutional symptoms. It also remains obtained human PBMCs, concentrations their function. Autophagy is also observed unclear whether HDACi’s will be used of 12.5 to 25 nmol/L of ITF2357 were in yeasts, and the weak HDACi valproic α as a monotherapy or in combination observed for IC50 of LPS-induced TNF , acid increases autophagy in yeasts under- with other antiinflammatory agents such IL-1β and IFNγ. going the stress of DNA breaks (121).

346 | DINARELLO ET AL. | MOL MED 17(5-6)333-352, MAY-JUNE 2011 REVIEW ARTICLE

Without this mechanism, cells die with in- tion after pulmonary artery banding in show the utility of these studies, the au- tracellular debris and organs fail. In some rats via a mechanism of decreased angio- thors treated mice with established left disease processes, autophagy can be mal- genesis and eNOS expression (125). ventricular hypertrophy with TSA (1.0 adaptive for the host, and this is certainly Whereas HDAC inhibition reduces left mg/kg/day) and observed that ventricu- the case in malignant cells. HDACi’s func- ventricular hypertrophy, fibrosis and apo- lar function returned to normal levels tion in the treatment of cancer by increas- ptosis, the opposite effects of HDACi (124). Similar protective results of ing genes that inhibit autophagy. treatment were observed with pressure HDACi’s were reported in animal models In diabetes, autophagy is a protective overload of the right ventricle (125). It is of ischemic and nonischemic (doxoru- process (122,123). Endogenous proteins unclear from that study of pulmonary bicin) myocardial injury (126,127). These that are trapped in the endoplasmic retic- banding–induced right ventricular failure studies and the role of HDACi’s in heart ulum (ER) result in ER stress. In the if lower doses of TSA would have yielded failure are reviewed by McKinsey (63) in insulin-producing pancreatic β cell, ER the same data in vivo. In vitro, 300 nmol/L this issue. The data suggest that oral stress can occur because of a failure to re- TSA was required to reduce eNOS and HDACi’s may have a role in treating left lease insulin into the extracellular com- VEGF in cultured human cardiac mi- ventricular hypertrophy, while caution partment after glucose load and die an crovessel endothelial cells, a concentration should be used in cases of right ventricu- β apoptotic death. Cells producing IL-1 100-fold greater for the IC50 of TSA for lar hypertrophy. such as the β cell itself have an additional HDAC-2 and 150-fold greater for HDAC-1 threat related to caspase-1, that is, the pro- (7). However, these authors are correct in HDACi’s AS A METHOD FOR HIV-1 cessing of the IL-1β precursor and release that HDACi’s do reduce angiogenesis, as ERADICATION of active IL-1β. Active IL-1β can then trig- we have observed blood vessel formation Since the introduction of highly active ger the IL-1 receptor type I (IL-1RI) lead- in Matrigel plugs embedded with 10 ng antiretroviral therapy (ART) more than a ing to NO production and cell death. IL-1β in mice treated with 2.5 mg/kg oral decade ago, HIV-1 infection can be well HDACi’s reduce IL-1β processing and re- ITF235 (unpublished data). controlled, and in nearly all treated pa- lease and therefore are of potential use in Left ventricular hypertrophy takes place tients, HIV-1 viremia remains below de- protecting the β cell and thus are a treat- as a compensatory mechanism to restore tectable levels as long as therapy is sus- ment possibility in diabetes. functionality in conditions of increased af- tained. Indeed, ART has changed the face terload or wall stress (that is, after an is- of HIV-1. For example, after the initiation HDACi’s and the Failing Heart chemic event). Left ventricular hypertro- of ART, markedly reduced CD4+ T cells Although autophagy benefits the phy is an unfavorable prognostic factor rose to near-normal levels, and death from insulin-producing β cell, autophagy is and represents a major clinical problem af- opportunistic infection was rarely ob- maladaptive in left-side ventricular hyper- fecting millions of patients worldwide. In- served. As such, this formally fatal disease trophic heart disease (124). In the stressed deed, the concept has been tested in has been transformed into a chronic dis- ischemic heart, the autophagic process at- mouse models of left ventricular hypertro- ease. Nevertheless, lifespan, even in pa- tempts to prevent loss of the myocyte and phy. The study used phenylephrine-in- tients well controlled on ART, remains reuse the intracellular debris generated duced hypertrophic growth and was pre- below that of the general population. De- during the ischemic event. Inhibition of vented by treatment with TSA (124). Other spite the effectiveness of ART in suppress- autophagy would seem counterproductive HDACi’s reduced hypertrophic growth in ing HIV-1 replication, after discontinua- for an ischemic heart. HDACi’s do, in fact, cultured cardiomyocytes. Hypertrophic tion of ART, there is a rebound viremia, inhibit autophagy, and therefore their use agonist–induced autophagy is suppressed which occurs typically after 2 weeks (128). would be counterproductive in rescuing in neonatal cardiomyocytes in culture The source of viral reemergence is a long- cells from death. However, in hyper- when treated with the hypertrophic ago- lived pool, most likely the latently infected trophic left ventricle heart failure, prevent- nists in the presence of TSA (66 nmol/L) memory CD4+ T-cell reservoir, harboring ing autophagy is beneficial, since it re- or valproic acid (6 mmol/L) (124). To integrated HIV-1 proviral DNA (129). At- duces the proliferation. In many cases of demonstrate the role of autophagy in the tacking this latent pool remains a priority heart failure, there is a constriction of the hypertrophic cardiomyocyte growth, ex- in combating HIV-1 infections (130). aortic outflow, and the hypertrophic left pression of endogenous ATF5, the gene The latent reservoir of HIV-1 within ventricle undergoes a proliferative re- that is essential for autophagic protection, resting CD4+ cells is established early sponse, which is necessary to compensate was suppressed with inhibitory RNA and after acute infection and is an extremely for the restriction. The benefit of HDACi’s resulted in reduced cardiomyocyte growth stable reservoir, having a half-life of in this form of heart failure is not the case in vitro. The ability of HDACi’s to inhibit 6–44 months, even in treated patients in right side heart failure due to pressure autophagy appeared to target HDAC-1 who are continuously aviremic for long overload. Suppression of histone deacety- and HDAC-2 (124); thus, HDAC-1 and -2 periods of time (129). The integrated lases worsens right ventricular dysfunc- contribute to maintaining autophagy. To proviral DNA is present in densely orga-

MOL MED 17(5-6)333-352, MAY-JUNE 2011 | DINARELLO ET AL. | 347 HDAC INHIBITORS IN NONCANCER DISEASES

nized nucleosomes; for viral replication, (vorinostat) is more potent than valproic duction in platelet count was observed transcription factors such as NFκB bind acid, has greater specificity for HDAC-1 across all doses. The effect was evident to the long terminal repeat of HIV-1 and and HDAC-2 and is in clinical trials in from day 5 onward, reaching the lowest drives expression of HIV-1 with release of HIV-1-infected patients while taking ART value on day 9, that is, 48 hours after last competent virus. During the process of (131,133). As reported in this issue of dosing. On day 9, the mean platelet viral replication, the CD4+ T cells die, but Molecular Medicine, ITF2357 (givinostat) counts decreased by 17%, 25% and 35% new CD4+ T cells are infected after the re- appears to be more potent and specific in groups receiving givinostat at 50, 100 lease of a competent virus. However, in than vorinostat (19,38). We believe givino- and 200 mg each day, respectively. For patients on ART, any newly released stat should be considered for a clinical the most part, the effect on platelets re- virus cannot be integrated into the DNA trial to force HIV-1 out of its latency and covered on day 14 (7 days after last dos- of naïve CD4+ T cells, since the activity of gradually eliminate the infection. ing). In children treated with givinostat reverse transcriptase and intergrases are The hypothesis assumes that, with ex- for 12 weeks at a dose of 1.5 mg/kg in inhibited by the antiretroviral drugs. The pression of HIV-1 from it latency, the in- two divided doses, there was a fall in use of the chemokine/HIV-1 receptor an- fected CD4+ T cell dies of a cytopathic platelets; however, disease activity in tagonists also prevents viral entry. To de- process and that the newly released virus SOJIA includes thromocytosis, and fall in plete the latent pool, activating the CD4+ is incapable of being integrated into platelets is interpreted as beneficial. Nev- T cell with IL-2 therapy in patients on naïve CD4+ T cells. As reported by ertheless, at a dose of 1.5 mg/kg, platelet ART was tested in large clinical trials but Matalon et al. (19) and reviewed in this count remained in the normal range. without a significant reduction. Another issue (38), givinostat induces HIV-1 from For givinostat, a dose between 1.0 and approach is to force viral expression, by- latently infected human macrophages 1.5 mg/kg in two divided doses is well passing activation of the CD4+ T cells. and T-cell lines at concentrations achiev- tolerated and unlikely to reduce platelets In latently infected CD4+ T cells, able in humans and at doses that are to below the normal range. If a particular HDAC-1 and HDAC-2 deacetylate local well tolerated. By comparison, to achieve HDACi is well tolerated in terms of not histones, compact the chromatin and pre- levels of vorinostat in vivo comparable to inducing constitutional symptoms related vent the binding of NFκB and other tran- the concentrations required to induce to gastrointestinal disturbances or de- scription factors, and RNA polymerase II HIV-1 expression in latently infected creasing daily functions, a fall in platelets binding does not occur. In vitro studies human macrophage and T-cell lines, the in the range of 15–20% is without clinical demonstrated that activation of a latent drug would likely not be well tolerated. significance. Nevertheless, regardless of viral by NFκB reverses the repressive ef- Similarly, to obtain the same level of the dosing schedule, any use of HDACi’s fect of the p50 homodimer HDAC-1 com- HIV-1 expression from these cell lines in for treating a chronic disease will require plex by the binding of cytosolic NFκB vitro, valproic acid concentrations in vivo monitoring of platelet counts. p50-RelA heterodimer. This step would would not be well tolerated. Similar to The fall in platelets is not due to bone enable the recruitment of the histone other treatment options of HDACi’s in marrow suppression. The HDACi ro- acetyltransferase, acetylation of the local chronic diseases, as summarized in this midepsin (FK228) is approved for the histones, relaxation of the chromatin and issue of Molecular Medicine, any disease- treatment of cutaneous T-cell lymphoma initiation of viral transcription. Inhibition modifying benefit of an HDACi must be and is selective for HDAC-1 and -2 of the enzymatic activity of HDAC-1 and at a dose that is well tolerated. (135). In examining the mechanism for possibly other HDACs by HDACi’s re- the reduction in platelets, it was con- sults in activation of the HIV-1 long ter- THROMBOCYTOPENIA cluded that romidepsin in mice is not minal repeat, and viral expression is ob- HDACi’s used to treat malignant dis- myelosuppressive and does not reduce served in latently infected cells in vitro. ease consistently induce thrombocyto- the lifespan of platelets (134). Rather, Because the HDACi valproic acid stim- penia. In fact, in addition to gastrointesti- studies with romidepsin demonstrated ulates HIV-1 expression in vitro, valproic nal disturbances such as nausea and that HDACi’s are more likely to cause a acid was tested in trials of HIV-1 in in- diarrhea, a low platelet count appears to reduction in platelets because of the im- fected patients while on continuous ART. be the major dose-limiting toxicity (134). paired release of platelets from the However, those trials did not yield the de- A fall in platelets has also been observed megakaryocytes. In the romidepsin sirable decrease in the latent reservoir in healthy subjects during a phase 1 study, primary megakaryocytes from (116,131,132). It is believed that valproic study of givinostat in healthy subjects mice showed reductions in proplatelet acid failed to have a significant impact be- (87). In that study, there was no concomi- extensions with an increase in the phos- cause it is a nonspecific and weak HDACi. tant use of chemotherapy or radiation, as phorylation of myosin light chain 2 Thus, whether the use of HDACi’s can is often the case with cancer patients re- (134). Of particular interest in that study purge the virus from the latent reservoir ceiving HDACi’s. In subjects treated with was the observation that that the requires more potent agents. SAHA daily dosing of givinostat for 7 days, a re- HDACi-induced thrombocytopenia in

348 | DINARELLO ET AL. | MOL MED 17(5-6)333-352, MAY-JUNE 2011 REVIEW ARTICLE

mice was prevented by coadministration onset juvenile idiopathic arthritis. Arthritis 23. Grabiec AM, et al. (2010) Histone deacetylase in- of a thrombopoietin mimetic (134). Rheum. 63:1452–8. hibitors suppress inflammatory activation of 6. Lewis EC, et al. (2011) The oral histone deacety- rheumatoid arthritis patient synovial macrophages and tissue. J. Immunol. 184:2718–28. SUMMARY lase inhibitor ITF2357 reduces cytokines and protects islet β cells in vivo and in vitro. Mol. Med. 24. Grabiec AM, Tak PP, Reedquist KA. (2011) His- Studies in animal models have re- 17:369–377. tone deacetylase inhibitors suppress IL-6 produc- vealed the unexpected observations that 7. Khan N, et al. (2008) Determination of the class tion by rheumatoid arthritis fibroblast-like syn- HDACi’s can be used to reduce severity and isoform selectivity of small-molecule histone oviocytes and macrophages via modulation of in a broad range of diseases. For the deacetylase inhibitors. Biochem. J. 409:581–9. mRNA stability rather than blockade of NFkB 8. Wang WC. (2008) The pharmacotherapy of sickle signalling. Ann. Rheum. Dis. 70 Suppl 2:A30–31. most part, the spectrum of diseases is in- cell disease. Expert Opin. Pharmacother. 9:3069–82. 25. Joosten LAB, Leoni F, Meghji S, Mascagni P. flammatory; and hence HDACi’s exhibit 9. Hines P, Dover GL, Resar LM. (2008) Pulsed-dos- (2011) Inhibition of HDAC activity by ITF2357 antiinflammatory properties owing to a ing with oral sodium phenylbutyrate increases ameliorates joint inflammation and prevents car- reduction in cytokine production as well hemoglobin F in a patient with sickle cell ane- tilage and bone destruction in experimental arthritis. Mol. Med. 17:391–396. as inhibition of cytokine effects. Distinct mia. Pediatric Blood Cancer. 50:357–59. 26. Vojinovic J, Damjanov N. (2011) HDAC inhibi- from their use in cancer, the effects of 10. Leoni F, et al. (2002) The antitumor histone deacetylase inhibitor suberoylanilide hydroxamic tion in rheumatoid arthritis and juvenile idio- HDACi’s are consistently observed at acid exhibits antiinflammatory properties via pathic arthritis. Mol. Med. 17:397–403. low concentrations compared to higher suppression of cytokines. Proc. Natl. Acad. Sci. 27. Chen CJ, et al. (2007) Identification of a key path- concentrations required for killing tumor U. S. A. 99:2995–3000. way required for the sterile inflammatory response 11. Leoni F, et al. (2005) The histone deacetylase in- triggered by dying cells. Nat. Med. 13:851–6. cells. HDACi’s are therefore attractive as 28. Cohen I, et al. (2010) Differential release of chro- hibitor ITF2357 reduces production of pro- a therapeutic option for suppressing matin-bound IL-1alpha discriminates between inflammatory cytokines in vitro and systemic in- necrotic and apoptotic cell death by the ability to inflammation. flammation in vivo. Mol. Med. 11:1–15. induce sterile inflammation. Proc. Natl. Acad. Sci. 12. Curtin M, Glaser K. (2003) Histone deacetylase U. S. A. 107:2574–9. ACKNOWLEDGMENTS inhibitors: the Abbott experience. Curr. Med. 29. Scaffidi P, Misteli T, Bianchi ME. (2002) Release Chem. 10:2373–92. CA Dinarello was supported by Na- of chromatin protein HMGB1 by necrotic cells 13. Montgomery RL, et al. (2007) Histone deacety- tional Institutes of Health Grant AI-15614. triggers inflammation. Nature 418:191–5. lases 1 and 2 redundantly regulate cardiac mor- The authors thank A. Grabiek and K. 30. Miossec P, Korn T, Kuchroo VK. (2009) Inter- phogenesis, growth, and contractility. Genes Dev. leukin-17 and type 17 helper T cells. N. Engl. J. Reedquist for insights into HDAC gene 21:1790–802. Med. 361:888–98. deletion studies and A. Abbate for valu- 14. Haberland M, Mokalled MH, Montgomery RL, 31. Donath MY, Storling J, Berchtold LA, Billestrup Olson EN. (2009) Epigenetic control of skull mor- able interpretation of differences between N, Mandrup-Poulsen T. (2008) Cytokines and phogenesis by histone deacetylase 8. Genes Dev. left and right ventricular heart failure. We beta-cell biology: from concept to clinical transla- 23:1625–30. also thank each of the contributors to the tion. Endocr. Rev. 29:334–50. 15. Knutson SK, et al. (2008) Liver-specific deletion 32. Larsen L, et al. (2007) Inhibition of histone 14 articles in this issue of Molecular Medi- of histone deacetylase 3 disrupts metabolic tran- deacetylases prevents cytokine-induced toxicity cine. They have provided readers with scriptional networks. EMBO J. 27:1017–28. in beta cells. Diabetologia. 50:779–89. their unique concepts on the mechanisms 16. Klampfer L, Huang J, Swaby LA, Augenlicht L. 33. Susick L, Veluthakal R, Suresh MV, Hadden T, for using HDACi’s in treating diseases in (2004) Requirement of histone deacetylase activity Kowluru A. (2007) Regulatory roles for histone for signaling by STAT1. J. Biol. Chem. 279:30358–68. their respective areas of expertise. deacetylation in IL-1beta-induced nitric oxide re- 17. Dinarello CA. (2011) Interleukin-1 in the patho- lease in pancreatic beta-cells. J. Cell Mol. Med. 5:5. genesis and treatment of inflammatory diseases. 34. Christensen DP, et al. (2011) Histone deacetylase DISCLOSURE Blood. 117:3720–32. (HDAC) inhibition as a novel treatment for dia- CA Dinarello is a consultant to 18. Hoffman HM, Wanderer AA. Inflammasome and betes mellitus. Mol. Med. 17:378–390. Italfarmaco. G Fossatiand P Mascagni IL-1beta-mediated disorders. Curr. Allergy 35. Lundh M, et al. (2010) Lysine deacetylases are are employees of Italfarmaco. Asthma Rep. 10:229–35. produced in pancreatic beta cells and are differ- 19. Matalon S, et al. (2010) The histone deacetylase entially regulated by proinflammatory cytokines. inhibitor ITF2357 decreases surface CXCR4 and Diabetologia. 53:2569–78. REFERENCES CCR5 expression on CD4(+) T-cells and mono- 36. Patel T, Patel V, Singh R, Jayaraman S. (2011) 1. Dinarello CA. (2010) Anti-inflammatory agents: cytes and is superior to valproic acid for latent Chromatin remodeling resets the immune system present and future. Cell. 140:935–50. HIV-1 expression in vitro. J. Acquir. Immune Defic. to protect against autoimmune diabetes in mice. 2. Mandrup-Poulsen T, Pickersgill L, Donath MY. Syndr. 54:1–9. Immunol. Cell Biol. 2011, Feb 15 [Epub ahead of (2010) Blockade of interleukin 1 in Type 1 dia- 20. Bosisio D, et al. (2008) Blocking TH17-polarizing print]. betes mellitus. Nat. Rev. Endocrinol. 6:158–66. cytokines by histone deacetylase inhibitors in 37. Koulmanda M, et al. (2008) Curative and beta cell 3. Donath MY, Shoelson SE. (2011) Type 2 diabetes vitro and in vivo. J. Leukoc. Biol. 84:1540–8. regenerative effects of alpha1-antitrypsin treat- as an inflammatory disease. Nat. Rev. Immunol. 21. Wang H, et al. (2011) Histone deacetylase in- ment in autoimmune diabetic NOD mice. Proc. 11:98–107. hibitor LAQ824 augments inflammatory re- Natl. Acad. Sci. U. S. A. 105:16242–7. 4. Dinarello CA, Donath MY, Mandrup-Poulsen T. sponses in macrophages through transcriptional 38. Matalon S, Rasmussen TA, Dinarello CA. (2011) (2010) Role of IL-1beta in Type 2 diabetes. Curr. regulation of IL-10. J. Immunol. 186:3986–96. Histone deacetylase inhibitors for purging HIV-1 Opin. Endocrinol. Diabetes Obes. 17:314–21. 22. Song W, et al. (2010) HDAC inhibition by LBH589 from the latent reservoir. Mol. Med. 17:466–472. 5. Vojinovic J, et al. (2011) Safety and efficacy of an affects the phenotype and function of human 39. Crazzolara R, et al. (2002) Histone deacetylase in- oral histone deacetylase inhibitor in systemic- myeloid dendritic cells. Leukemia. 25:161–8. hibitors potently repress CXCR4 chemokine re-

MOL MED 17(5-6)333-352, MAY-JUNE 2011 | DINARELLO ET AL. | 349 HDAC INHIBITORS IN NONCANCER DISEASES

ceptor expression and function in acute lym- hibitors impair innate immune responses to Toll- tone deacetylases in inflammatory bowel dis- phoblastic leukaemia. Br. J. Haematol. 119:965–9. like receptor agonists and to infection. Blood. eases. Mol. Med. 17:426–433. 40. Larsen CM, et al. (2007) Interleukin-1-receptor 117:1205–17. 75. Reilly CM, Regna N, Mishra N. (2011) HDAC in- antagonist in Type 2 diabetes mellitus. N. Engl. J. 58. Ramirez-Carrozzi VR, et al. (2006) Selective and hibition in lupus models. Mol. Med. 17:417–425. Med. 356:1517–26. antagonistic functions of SWI/SNF and Mi-2beta 76. Faraco G, Cavone L, Chiarugi A. (2011) The ther- 41. Faraco G, et al. (2009) Histone deacetylase (HDAC) nucleosome remodeling complexes during an in- apeutic potential of HDAC inhibitors in the treat- inhibitors reduce the glial inflammatory response flammatory response. Genes Dev. 20:282–96. ment of multiple sclerosis. Mol. Med. 17:442–447. in vitro and in vivo. Neurobiol. Dis. 36:269–79. 59. Calabrese F, et al. (2008) IL-32, a novel proinflam- 77. Shein NA, Shohami E. (2011) Histone deacetylase 42. Green SR, Choudhary AK, Fleming IN. (2009) matory cytokine in chronic obstructive pulmonary inhibitors as therapeutic agents for acute central Combination of sapacitabine and HDAC in- disease. Am. J. Respir. Crit. Care Med. 178:894–901. nervous system injuries. Mol. Med. 17:448–456. hibitors stimulates cell death in AML and other 60. Ito K, et al. (2005) Decreased histone deacetylase 78. Consalvi S, et al. (2011) Histone deacetylase in- tumour types. Br. J. Cancer. 103:1391–9. activity in chronic obstructive pulmonary dis- hibitors in the treatment of muscular dystro- 43. Choi S, Reddy P. (2011) HDAC inhibition and ease. N. Engl. J. Med. 352:1967–76. phies: epigenetic drugs for genetic diseases. Mol. graft versus host disease. Mol. Med. 17:404–416. 61. Mizuno S, et al. (2010) Inhibition of histone Med. 17:457–465. 44. Reddy P, et al. (2008) Histone deacetylase inhibi- deacetylase causes emphysema. Am. J. Physiol. 79. Halili MA, Andrews MR, Sweet MJ, Fairlie DP. tion modulates indoleamine 2,3-dioxygenase- Lung Cell Mol. Physiol. 300:L402–13. (2009) Histone deacetylase inhibitors in inflam- dependent DC functions and regulates experi- 62. Dinarello CA. (2005) Differences between anti- matory disease. Curr. Top. Med. Chem. 9:309–19. mental graft-versus-host disease in mice. J. Clin. tumor necrosis factor-alpha monoclonal antibod- 80. Choi JH, et al. (2005) Trichostatin A exacerbates Invest. 118:2562–73. ies and soluble TNF receptors in host defense im- atherosclerosis in low density lipoprotein recep- 45. Choudhary C, et al. (2009) Lysine acetylation tar- pairment. J. Rheumatol. Suppl. 74:40–7. tor-deficient mice. Arterioscler. Thromb. Vasc. Biol. gets protein complexes and co-regulates major 63. McKinsey TA. (2011) Targeting inflammation in 25:2404–9. cellular functions. Science. 325:834–40. heart failure with histone deacetylase inhibitors. 81. Halili MA, et al. (2010) Differential effects of se- 46. Calao M, Burny A, Quivy V, Dekoninck A, Van Mol. Med. 17:434–441. lective HDAC inhibitors on macrophage inflam- Lint C. (2008) A pervasive role of histone acetyl- 64. Abbate A, et al. (2010) Interleukin-1beta modula- matory responses to the Toll-like receptor 4 ago- transferases and deacetylases in an NF-kappaB- tion using a genetically engineered antibody pre- nist LPS. J. Leukoc. Biol. 87:1103–14. signaling code. Trends Biochem. Sci. 33:339–49. vents adverse cardiac remodelling following 82. Zayed N, et al. (2008) Inhibition of interleukin- 47. Chen LF, Mu Y, Greene WC. (2002) Acetylation of acute myocardial infarction in the mouse. Eur. J. 1beta-induced matrix metalloproteinases 1 and 13 RelA at discrete sites regulates distinct nuclear Heart Fail. 12:319–22. production in human osteoarthritic chondrocytes functions of NF-kappaB. EMBO J. 21:6539–48. 65. Abbate A, et al. (2010) Interleukin-1 blockade by prostaglandin D2. Arthritis Rheum. 58:3530–40. 48. Chen L, Fischle W, Verdin E, Greene WC. (2001) with anakinra to prevent adverse cardiac remod- 83. Yamaguchi K, Lantowski A, Dannenberg AJ, Sub- Duration of nuclear NF-kappaB action regulated eling after acute myocardial infarction. Am. J. baramaiah K. (2005) Histone deacetylase inhibitors by reversible acetylation. Science. 293:1653–7. Cardiol. 105:1371–7. suppress the induction of c-Jun and its target 49. Kiernan R, et al. (2003) Post-activation turn-off of 66. Deswal A, et al. (1999) Safety and efficacy of a genes including COX-2. J. Biol. Chem. 280:32569–77. NF-kappa B-dependent transcription is regulated soluble P75 tumor necrosis factor receptor (En- 84. Aung HT, et al. (2006) LPS regulates proinflam- by acetylation of p65. J. Biol. Chem. 278:2758–66. brel, etanercept) in patients with advanced heart matory gene expression in macrophages by alter- 50. Bode KA, et al. (2007) Histone deacetylase in- failure. Circulation. 99:3224–6. ing histone deacetylase expression. FASEB J. hibitors decrease Toll-like receptor-mediated acti- 67. Mann DL. (2002) Inflammatory mediators and 20:1315–27. vation of proinflammatory gene expression by the failing heart: past, present, and the foresee- 85. Garcia-Manero G, et al. (2008) Phase 1 study of impairing transcription factor recruitment. Immu- able future. Circ. Res. 91:988–98. the histone deacetylase inhibitor vorinostat nology. 122:596–606. 68. Di Iorio A, et al. (2003) Serum IL-1beta levels in (suberoylanilide hydroxamic acid [SAHA]) in pa- 51. Cao W, Bao C, Padalko E, Lowenstein CJ. (2008) health and disease: a population-based study: tients with advanced leukemias and myelodys- Acetylation of mitogen-activated protein kinase “The InCHIANTI study.” Cytokine. 22:198–205. plastic syndromes. Blood. 111:1060–6. phosphatase-1 inhibits Toll-like receptor signal- 69. Abbate A, et al. (2008) Anakinra, a recombinant 86. O’Connor OA, et al. (2006) Clinical experience with ing. J. Exp. Med. 205:1491–503. human interleukin-1 receptor antagonist, inhibits intravenous and oral formulations of the novel his- 52. Carta S, et al. (2006) Histone deacetylase in- apoptosis in experimental acute myocardial in- tone deacetylase inhibitor suberoylanilide hydrox- hibitors prevent exocytosis of interleukin-1beta- farction. Circulation. 117:2670–83. amic acid in patients with advanced hematologic containing secretory lysosomes: role of micro- 70. Joosten LA, et al. (1999) IL-1 alpha beta blockade malignancies. J. Clin. Oncol. 24:166–73. tubules. Blood. 108:1618–26. prevents cartilage and bone destruction in 87. Furlan A, et al. (2011) Pharmacokinetics, safety 53. Martinon F, Mayor A, Tschopp J. (2009) The in- murine type II collagen-induced arthritis, and inducible cytokine responses during a phase flammasomes: guardians of the body. Annu. Rev. whereas TNF-alpha blockade only ameliorates 1 trial of the oral histone deacetylase inhibitor Immunol. 27:229–65. joint inflammation. J. Immunol. 163:5049–55. ITF2357 (givinostat). Mol. Med. 17:353–362. 54. Into T, Inomata M, Niida S, Murakami Y, Shibata 71. Reddy P, et al. (2004) Histone deacetylase in- 88. Lin T, et al. (2006) Cardiac histones are substrates K. (2010) Regulation of MyD88 aggregation and hibitor suberoylanilide hydroxamic acid reduces of histone deacetylase activity in hemorrhagic the MyD88-dependent signaling pathway by se- acute graft-versus-host disease and preserves shock and resuscitation. Surgery. 139:365–76. questosome 1 and histone deacetylase 6. J. Biol. graft-versus-leukemia effect. Proc. Natl. Acad. Sci. 89. Sailhamer EA, et al. (2008) Acetylation: a novel Chem. 285:35759–69. U. S. A. 101:3921–6. method for modulation of the immune response 55. Li Y, et al. (2010) Surviving lethal septic shock 72. Sun Y, et al. (2009) Cutting edge: negative regulation following trauma/hemorrhage and inflamma- without fluid resuscitation in a rodent model. of dendritic cells through acetylation of the nonhis- tory second hit in animals and humans. Surgery. Surgery. 148:246–54. tone protein STAT-3. J. Immunol. 182:5899–903. 144:204–16. 56. Brogdon JL, et al. (2007) Histone deacetylase ac- 73. Bodar EJ, Simon A, van der Meer JWM. (2011) Ef- 90. Terkeltaub R, et al. (2009) The interleukin 1 in- tivities are required for innate immune cell con- fects of the histone deacetylase inhibitor ITF2357 hibitor rilonacept in treatment of chronic gouty trol of Th1 but not Th2 effector cell function. in autoinflammatory syndromes. Mol. Med. arthritis: results of a placebo-controlled, monose- Blood. 109:1123–30. 17:363–368. quence crossover, non-randomised, single-blind 57. Roger T, et al. (2011) Histone deacetylase in- 74. Glauben R, Siegmund B. (2011) Inhibition of his- pilot study. Ann. Rheum. Dis. 68:1613–7.

350 | DINARELLO ET AL. | MOL MED 17(5-6)333-352, MAY-JUNE 2011 REVIEW ARTICLE

91. So A, De Smedt T, Revaz S, Tschopp J. (2007) A myeloma. Hematol. Oncol. Clin. North Am. 122. Hur KY, Jung HS, Lee MS. (2010) Role of au- pilot study of IL-1 inhibition by anakinra in 13:1117–25. tophagy in beta-cell function and mass. Diabetes acute gout. Arthritis Res. Ther. 9:R28. 107. Galli M, et al. (2010) A phase II multiple dose Obes. Metab. 12 Suppl 2:20–26. 92. So A, et al. (2010) Canakinumab for the treat- clinical trial of histone deacetylase inhibitor 123. Vanhorebeek I, et al. (2011) Insufficient activation ment of acute flares in difficult-to-treat gouty ITF2357 in patients with relapsed or progressive of autophagy allows cellular damage to accu- arthritis: results of a multicenter, phase II, dose- multiple myeloma. Ann. Hematol. 89:185–90. mulate in critically ill patients. J. Clin. Endocrinol. ranging study. Arthritis Rheum. 62:3064–76. 108. Todoerti K, et al. (2010) Pleiotropic anti-myeloma Metab. 96:E633-45. 93. Shein NA, et al. (2009) Histone deacetylase in- activity of ITF2357: inhibition of interleukin-6 124. Cao DJ, et al. (2011) Histone deacetylase (HDAC) hibitor ITF2357 is neuroprotective, improves receptor signaling and repression of miR-19a inhibitors attenuate cardiac hypertrophy by sup- functional recovery, and induces glial apoptosis and miR-19b. Haematologica. 95:260–9. pressing autophagy. Proc. Natl. Acad. Sci. U. S. A. following experimental traumatic brain injury. 109. Rambaldi A, et al. (2010) A pilot study of the his- 108:4123–8. FASEB J. 23:4266–75. tone-deacetylase inhibitor givinostat in patients 125. Bogaard HJ, et al. (2011) Suppression of histone 94. Kozikowski AP, et al. (2009) Searching for disease with JAK2V617F positive chronic myeloprolifer- deacetylases worsens right ventricular dysfunc- modifiers-PKC activation and HDAC inhibition: ative neoplasms. Br. J. Haematol. 150:446–55. tion after pulmonary artery banding in rats. a dual drug approach to Alzheimer’s disease 110. Guerini V, et al. (2008) The histone deacetylase Am. J. Respir. Crit. Care Med. 2011, Feb 4 [Epub that decreases Abeta production while blocking inhibitor ITF2357 selectively targets cells bear- ahead of print]. oxidative stress. Chem. Med. Chem. 4:1095–105. ing mutated JAK2(V617F). Leukemia. 22:740–7. 126. Daosukho C, et al. (2007) Phenylbutyrate, a his- 95. Ona VO, et al. (1999) Inhibition of caspase-1 111. Hansen ER, Vejlsgaard GL, Lisby S, Heidenheim tone deacetylase inhibitor, protects against slows disease progression in a mouse model of M, Baadsgaard O. (1991) Epidermal interleukin Adriamycin-induced cardiac injury. Free Radic. Huntington’s disease. Nature 399:263–7. 1 alpha functional activity and interleukin 8 im- Biol. Med. 42:1818–25. 96. Hockly E, et al. (2003) Suberoylanilide hydroxamic munoreactivity are increased in patients with 127. Granger A, et al. (2008) Histone deacetylase in- acid, a histone deacetylase inhibitor, ameliorates cutaneous T-cell lymphoma. J. Invest. Dermatol. hibition reduces myocardial ischemia-reperfu- motor deficits in a mouse model of Huntington’s 97:818–23. sion injury in mice. FASEB J. 22:3549–60. disease. Proc. Natl. Acad. Sci. U. S. A. 100:2041–6. 112. Bladon J, Taylor PC. (2006) The down-regulation 128. Davey RT Jr, et al. (1999) HIV-1 and T cell dy- 97. Benn CL, et al. (2009) Genetic knock-down of of IL1alpha and IL6, in monocytes exposed to namics after interruption of highly active anti- HDAC7 does not ameliorate disease pathogene- extracorporeal photopheresis (ECP)-treated lym- retroviral therapy (HAART) in patients with a sis in the R6/2 mouse model of Huntington’s phocytes, is not dependent on lymphocyte phos- history of sustained viral suppression. Proc. disease. PLoS One. 4:e5747. phatidylserine externalization. Transpl. Int. Natl. Acad. Sci. U. S. A. 96:15109–14. 98. Chuang DM. (2005) The antiapoptotic actions of 19:319–24. 129. Chun TW, et al. (2005) HIV-infected individuals mood stabilizers: molecular mechanisms and 113. Tilg H, et al. (1993) Induction of circulating IL-1 receiving effective antiviral therapy for ex- therapeutic potentials. Ann. N. Y. Acad. Sci. receptor antagonist by IFN treatment. J. Im- tended periods of time continually replenish 1053:195–204. munol. 150:4687–92. their viral reservoir. J. Clin. Invest. 115:3250–5. 99. Dompierre JP, et al. (2007) Histone deacetylase 6 114. Gerstner T, Bell N, Konig S. (2008) Oral valproic 130. Richman DD, et al. (2009) The challenge of find- inhibition compensates for the transport deficit acid for epilepsy: long-term experience in ther- ing a cure for HIV infection. Science. 323:1304–7. in Huntington’s disease by increasing tubulin apy and side effects. Expert Opin. Pharmacother. 131. Archin NM, et al. (2010) Antiretroviral intensifi- acetylation. J. Neurosci. 27:3571–83. 9:285–92. cation and valproic acid lack sustained effect on 100. Meissner F, Molawi K, Zychlinsky A. (2010) 115. Ren M, Leng Y, Jeong M, Leeds PR, Chuang residual HIV-1 viremia or resting CD4+ cell in- Mutant superoxide dismutase 1-induced IL- DM. (2004) Valproic acid reduces brain damage fection. PLoS One. 5:e9390. 1beta accelerates ALS pathogenesis. Proc. Natl. induced by transient focal cerebral ischemia in 132. Archin NM, et al. (2009) Expression of latent Acad. Sci. U. S. A. 107:13046–50. rats: potential roles of histone deacetylase inhi- human immunodeficiency type 1 is induced by 101. van der Meer JW, Simon A. (2010) Blocking IL- bition and heat shock protein induction. J. Neu- novel and selective histone deacetylase in- 1beta to slow down progression of ALS? Proc. rochem. 89:1358–67. hibitors. AIDS. 23:1799–806. Natl. Acad. Sci. U. S. A. 107:12741–2. 116. Archin NM, et al. (2008) Valproic acid without 133. Archin NM, et al. (2009) Expression of latent 102. Lust JA, et al. (2009) Induction of a chronic dis- intensified antiviral therapy has limited impact HIV induced by the potent HDAC inhibitor ease state in patients with smoldering or indo- on persistent HIV infection of resting CD4+ T suberoylanilide hydroxamic acid. AIDS Res. lent multiple myeloma by targeting interleukin cells. AIDS. 22:1131–5. Hum. Retroviruses. 25:207–12. 1{beta}-induced interleukin 6 production and 117. Atweh GF, Schechter AN. (2001) Pharmacologic 134. Bishton MJ, et al. (2011) Deciphering the molec- the myeloma proliferative component. Mayo induction of fetal hemoglobin: raising the thera- ular and biological processes that mediate his- Clin. Proc. 84:114–22. peutic bar in sickle cell disease. Curr. Opin. tone deacetylase inhibitor-induced thrombocy- 103. Dinarello CA. (2009) Targeting the pathogenic Hematol. 8:123–30. topenia. Blood. 117:3658–68. role of interleukin 1beta in the progression of 118. Bodar EJ, Simon A, van der Meer JWM. (2011) 135. Whittaker SJ, et al. (2010) Final results from a smoldering/indolent myeloma to active dis- Effects of the histone deacetylase inhibitor multicenter, international, pivotal study of ro- ease. Mayo Clin. Proc. 84:105–7. ITF2357 in autoinflammatory syndromes. Mol. midepsin in refractory cutaneous T-cell lym- 104. Golay J, et al. (2007) The histone deacetylase in- Med. 17:363–368. phoma. J. Clin. Oncol. 28:4485–91. hibitor ITF2357 has anti-leukemic activity in vitro 119. Mishra N, Reilly CM, Brown DR, Ruiz P, Gilke- 136. Choi Y, et al. (2008) Histone deacetylase in- and in vivo and inhibits IL-6 and VEGF produc- son GS. (2003) Histone deacetylase inhibitors hibitor KBH-A42 inhibits cytokine production tion by stromal cells. Leukemia. 21:1892–900. modulate renal disease in the MRL-lpr/lpr in RAW 264.7 macrophage cells and in vivo en- 105. Rambaldi A, et al. (1991) Modulation of cell pro- mouse. J. Clin. Invest. 111:539–52. dotoxemia model. Exp. Mol. Med. 40:574–81. liferation and cytokine production in acute 120. Reilly CM, et al. (2004) Modulation of renal dis- 137. Inoue K, et al. (2006) Histone deacetylase in- myeloblastic leukemia by interleukin-1 receptor ease in MRL/lpr mice by suberoylanilide hy- hibitor reduces monocyte adhesion to endothe- antagonist and lack of its expression by droxamic acid. J. Immunol. 173:4171–8. lium through the suppression of vascular cell leukemic cells. Blood. 78:3248–53. 121. Robert T, et al. (2011) HDACs link the DNA adhesion molecule-1 expression. Arterioscler. 106. Lust JA, Donovan KA. (1999) The role of inter- damage response, processing of double-strand Thromb. Vasc. Biol. 26:2652–9. leukin-1 beta in the pathogenesis of multiple breaks and autophagy. Nature. 471:74–9. 138. Su RC, Becker AB, Kozyrskyj AL, Hayglass KT.

MOL MED 17(5-6)333-352, MAY-JUNE 2011 | DINARELLO ET AL. | 351 HDAC INHIBITORS IN NONCANCER DISEASES

(2008) Epigenetic regulation of established FR235222-mediated IL-2 transcriptional repres- model of stroke: multiple mechanisms of action. human type 1 versus type 2 cytokine responses. sion in Jurkat cells. Int. Immunopharmacol. J. Pharmacol. Exp. Ther. 321:892–901. J. Allergy Clin. Immunol. 121:57–63.e3. 7:1422–32. 169. Anne-Laurence B, Caroline R, Irina P, Jean- 139. Choi JC, Holtz R, Murphy SP. (2009) Histone 155. Matsuoka H, et al. (2007) Disruption of Philippe L. (2007) Chromatin acetylation status deacetylases inhibit IFN-gamma-inducible gene HDAC4/N-CoR complex by histone deacety- in the manifestation of neurodegenerative dis- expression in mouse trophoblast cells. J. Im- lase inhibitors leads to inhibition of IL-2 gene eases: HDAC inhibitors as therapeutic tools. munol. 182:6307–15. expression. Biochem. Pharmacol. 74:465–76. Subcell. Biochem. 41:263–93. 140. Chabane N, et al. (2008) Histone deacetylase in- 156. Moreira JM, Scheipers P, Sorensen P. (2003) The 170. Janssen C, et al. (2010) Differential histone hibitors suppress interleukin-1beta-induced ni- histone deacetylase inhibitor trichostatin A deacetylase mRNA expression patterns in amy- tric oxide and prostaglandin E2 production in modulates CD4+ T cell responses. BMC Cancer. otrophic lateral sclerosis. J. Neuropathol. Exp. human chondrocytes. Osteoarthritis Cartilage. 3:30. Neurol. 69:573–81. 16:1267–74. 157. Skov S, et al. (2003) Histone deacetylase in- 171. Susick L, Senanayake T, Veluthakal R, Woster PM, 141. Crosson CE, Mani SK, Husain S, Alsarraf O, hibitors: a new class of immunosuppressors tar- Kowluru A. (2009) A novel histone deacetylase in- Menick DR. (2010) Inhibition of histone geting a novel signal pathway essential for hibitor prevents IL-1beta induced metabolic dys- deacetylase protects the retina from ischemic in- CD154 expression. Blood. 101:1430–8. function in pancreatic beta-cells. J. Cell Mol. Med. jury. Invest. Ophthalmol. Vis. Sci. 51:3639–45. 158. Lin HS, et al. (2007) Anti-rheumatic activities of 13:1877–85. 142. Iwata K, et al. (2002) Trichostatin A, a histone histone deacetylase (HDAC) inhibitors in vivo 172. Kinugasa F, et al. (2010) Prevention of renal in- deacetylase inhibitor, down-regulates inter- in collagen-induced arthritis in rodents. Br. J. terstitial fibrosis via histone deacetylase inhibi- leukin-12 transcription in SV-40-transformed Pharmacol. 150:862–72. tion in rats with unilateral ureteral obstruction. lung epithelial cells. Cell Immunol. 218:26–33. 159. Saouaf SJ, et al. (2009) Deacetylase inhibition in- Transpl. Immunol. 23:18–23. 143. Glauben R, et al. (2008) Histone deacetylases: creases regulatory T cell function and decreases 173. Marumo T, Hishikawa K, Yoshikawa M, Fujita novel targets for prevention of colitis-associated incidence and severity of collagen-induced T. (2008) Epigenetic regulation of BMP7 in the cancer in mice. Gut. 57:613–22. arthritis. Exp. Mol. Pathol. 87:99–104. regenerative response to ischemia. J. Am. Soc. 144. Glauben R, et al. (2006) Histone hyperacetyla- 160. Choo QY, Ho PC, Tanaka Y, Lin HS. (2010) His- Nephrol. 19:1311–20. tion is associated with amelioration of experi- tone deacetylase inhibitors MS-275 and SAHA 174. Marumo T, et al. (2011) Histone deacetylase mental colitis in mice. J. Immunol. 176:5015–22. induced growth arrest and suppressed modulates the proinflammatory and -fibrotic 145. de Zoeten EF, Wang L, Sai H, Dillmann WH, lipopolysaccharide-stimulated NF-kappaB p65 changes in tubulointerstitial injury. Am. J. Phys- Hancock WW. (2010) Inhibition of HDAC9 in- nuclear accumulation in human rheumatoid iol. Renal Physiol. 298:F133–41. creases T regulatory cell function and prevents arthritis synovial fibroblastic E11 cells. Rheuma- 175. Mie Lee Y, et al. (2003) Inhibition of hypoxia- colitis in mice. Gastroenterology. 138:583–94. tology (Oxford). 49:1447–60. induced angiogenesis by FK228, a specific his- 146. Tong X, Yin L, Giardina C. (2004) Butyrate sup- 161. Guo W, Shan B, Klingsberg RC, Qin X, Lasky tone deacetylase inhibitor, via suppression of presses Cox-2 activation in colon cancer cells JA. (2009) Abrogation of TGF-beta1-induced fi- HIF-1alpha activity. Biochem. Biophys. Res. Com- through HDAC inhibition. Biochem. Biophys. broblast-myofibroblast differentiation by his- mun. 300:241–6. Res. Commun. 317:463–71. tone deacetylase inhibition. Am. J. Physiol. Lung 176. Noh H, et al. (2009) Histone deacetylase-2 is a 147. Leng C, et al. (2006) Reduction of graft-versus- Cell Mol. Physiol. 297:L864–70. key regulator of diabetes- and transforming host disease by histone deacetylase inhibitor 162. Wang X, Song Y, Jacobi JL, Tuan RS. (2009) Inhi- growth factor-beta1-induced renal injury. Am. J. suberonylanilide hydroxamic acid is associated bition of histone deacetylases antagonized FGF2 Physiol. Renal Physiol. 297:F729–39. with modulation of inflammatory cytokine mi- and IL-1beta effects on MMP expression in 177. Yoshikawa M, Hishikawa K, Marumo T, Fujita lieu and involves inhibition of STAT1. Exp. human articular chondrocytes. Growth Factors. T. (2007) Inhibition of histone deacetylase activ- Hematol. 34:776–87. 27:40–9. ity suppresses epithelial-to-mesenchymal tran- 148. Tao R, et al. (2007) Deacetylase inhibition pro- 163. Kook H, et al. (2003) Cardiac hypertrophy and sition induced by TGF-beta1 in human renal ep- motes the generation and function of regulatory histone deacetylase-dependent transcriptional ithelial cells. J. Am. Soc. Nephrol. 18:58–65. T cells. Nat. Med. 13:1299–307. repression mediated by the atypical home- 149. Wang L, Tao R, Hancock WW. (2009) Using his- odomain protein Hop. J. Clin. Invest. 112:863–71. tone deacetylase inhibitors to enhance Foxp3(+) 164. Zhang CL, et al. (2002) Class II histone deacety- regulatory T-cell function and induce allograft lases act as signal-responsive repressors of car- tolerance. Immunol. Cell Biol. 87:195–202. diac hypertrophy. Cell. 110:479–88. 150. Edens RE, Dagtas S, Gilbert KM. (2006) Histone 165. Colussi C, et al. (2011) The histone deacetylase deacetylase inhibitors induce antigen specific inhibitor suberoylanilide hydroxamic acid re- anergy in lymphocytes: a comparative study. duces cardiac arrhythmias in dystrophic mice. Int. Immunopharmacol. 6:1673–81. Cardiovasc. Res. 87:73–82. 151. Tao R, et al. (2007) Histone deacetylase in- 166. Chen PS, et al. (2007) Valproic acid and other hi- hibitors and transplantation. Curr. Opin. Im- stone deacetylase inhibitors induce microglial munol. 19:589–95. apoptosis and attenuate lipopolysaccharide-in- 152. Camelo S, et al. (2005) Transcriptional therapy duced dopaminergic neurotoxicity. Neuroscience. with the histone deacetylase inhibitor tricho- 149:203–12. statin A ameliorates experimental autoimmune 167. Faraco G, et al. (2006) Pharmacological inhibi- encephalomyelitis. J. Neuroimmunol. 164:10–21. tion of histone deacetylases by suberoylanilide 153. Jung ID, et al. (2009) Apicidin, the histone hydroxamic acid specifically alters gene expres- deacetylase inhibitor, suppresses Th1 polariza- sion and reduces ischemic injury in the mouse tion of murine bone marrow-derived dendritic brain. Mol. Pharmacol. 70:1876–84. cells. Int. J. Immunopathol. Pharmacol. 22:501–15. 168. Kim HJ, et al. (2007) Histone deacetylase in- 154. Matsuoka H, Fujimura T, Mori H, Aramori I, hibitors exhibit anti-inflammatory and neuro- Mutoh S. (2007) Mechanism of HDAC inhibitor protective effects in a rat permanent ischemic

352 | DINARELLO ET AL. | MOL MED 17(5-6)333-352, MAY-JUNE 2011