Ann Rheum Dis 71:i46-i54 doi:10.1136/annrheumdis-2011-200593
  • Supplement

HDAC inhibitor therapy in autoimmunity and transplantation

  1. Liqing Wang1
  1. 1Department of Pathology and Laboratory Medicine, Children's Hospital of Philadelphia and University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, USA
  2. 2Division of Nephrology, Department of Pediatrics, Children's Hospital of Philadelphia and University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, USA
  1. Correspondence to Wayne W Hancock, Division of Transplant Immunology, Pathology and Laboratory Medicine, 916B Abramson Research Center, The Children's Hospital of Philadelphia, 3615 Civic Center Boulevard, Philadelphia, PA 19104-4318, USA; whancock{at}
  • Received 6 October 2011
  • Accepted 14 October 2011


Pharmacological inhibitors of histone/protein deacetylases (HDACi) have considerable therapeutic potential as anti-inflammatory and immunosuppressive drugs. The utility of HDACi as anti-inflammatory agents is dependent upon their proving safe and effective in experimental models. Current pan-HDACi compounds are ill-suited to this role, given the broad distribution of target HDACs and their complex and multifaceted mechanisms of action. By contrast, the development of isoform-selective HDACi may provide important new tools for treatment in autoimmunity and transplantation. This review discusses which HDACs are worthwhile targets in inflammation, and the progress made towards their therapeutic inhibition, including the use of HDAC subclass and isoform-selective HDACi to promote the functions of Foxp3+ T-regulatory cells.

Overview of HDACi

The fields of histone/protein deacetylase (HDAC) biology and HDAC targeting using small molecule inhibitors (HDACi) are intimately linked. Thus, the first HDACi compounds were identified by their anticancer effects during drug screens, before the existence of HDACs was known,1 and the HDACi, trapoxin, was used as a tool to isolate the first identified HDAC, HDAC1, in 1996.2 These inter-relationships continue to prove of utmost value in dissecting the roles of HDACs in cells and disease models. There are three broad classes of classical Zn-dependent HDACs, comprising class I, class II and class IV HDACs, and a single NAD-dependent sirtuin family of class III HDACs (SIRT1-7). Class I HDACs (HDAC1, -2, -3 and -8) are homologous to yeast Rpd3, class IIa (HDAC4, -5, -7 and -9) and class IIb (HDAC6 and -10) HDACs are homologous to yeast Hda1, and the sole class IV HDAC (HDAC11) has homology to both Rpd3 and Hda1.

Classic HDACs (HDAC1–11) have a cavity containing a catalytic Zn2+ ion that is connected to the protein–solvent interface by an 11 å tunnel.3 Classic HDACi come in various shapes and sizes, but a typical HDACi consists of a moiety that chelates the Zn2+ ion, has a variably sized linker to extend out from the tunnel, and one or more aryl groups that form a cap and interact with residues near the entrance to the binding pocket (figure 1). These residues have less homology between HDAC isoforms than those of the active site, allowing for different cap modifications to achieve relative HDAC isoform selectivity. Structure–activity relationship analysis has typically involved variations in the zinc-binding group and modifications of the cap-binding moiety, with the latter in some cases resulting in HDAC6-selective inhibitors (discussed below). The main classes of HDACi are hydroxamic acids (eg, trichostatin A (TsA), suberoylanilide hydroxamic acid (SAHA)), benzamides (eg, MS275), electrophilic ketones (eg, trapoxin), cyclic tetrapeptides (eg, depsipeptide (FK228), apicidin) and short-chain fatty acids (eg, butyrate, valproic acid).

Figure 1

Structural basis of HDAC inhibition. Diagram of the tunnel and active site of a classical HDAC, based on Wang et al,3 plus structures of two closely-related pan-HDACi hydroxamate compounds, TsA and SAHA, and an HDAC6 isoform-selective inhibitor, tubacin. This last compound, which also has an hydroxamic acid backbone, is thought to achieve its specificity as a result of its interactions with additional residues in the cap region. HDACi, histone/protein deacetylase inhibitors; SAHA, suberoylanilide hydroxamic acid; TsA, trichostatin A.

Despite the focus on modification of acetylation of histone tails suggested by the very names of these enzymes, much recent biology points to HDAC regulation of the levels of acetylation of non-histone proteins. Hence, HDACi can act in the cytoplasm (in the case of HDAC6) or the nucleus and modulate the functions of a large number of proteins. This is of more than of academic interest owing to rapid advances in the field, including recognition of an ‘acetylome’ probably encompassing thousands of proteins and responsible for the fine-regulation of functions in many normal cells.4 Moreover, while all HDACi in current clinical trials inhibit proliferation of transformed cells in culture by inducing cell cycle arrest and/or apoptosis, and inhibit tumour growth in animal models, far less is known about the effects of HDACi on the immune system. HDACi are known to induce lymphocyte cell cycle arrest, differentiation or apoptosis in vitro,5,,7 but in microarray analyses, the potent pan-HDACi, TsA, was found to modulate expression of only 2% of genes in T cell receptor-activated CD4+ T cells,7 comparable to studies showing that only 1–2% of genes were modified in non-T cells by HDACi.8

9 The effects of HDACi on protein acetylation and function would not be apparent from such mRNA-based studies, indicating a prime opportunity for new investigational studies and therapeutic applications. This review will summarise current data as to the therapeutic applications of HDACi in inflammation and autoimmunity, including the key cellular targets of HDACi, associated mechanisms of action, data from clinical studies and animal models of disease, and the rationale for therapeutic development of HDAC isoform selective agents for use in inflammation and autoimmunity. Further details are available from recent reviews on this topic.10,,17

Cells and diseases targeted by HDACi

Systemic administration of HDACi can potentially affect many, if not all, cells of the body depending upon whether a given agent can cross the blood–brain barrier. Some HDACi, such as valproic acid, do cross the blood–brain barrier and achieve high levels in the brain,18 whereas at normal dosing, hydroxamates, such as TsA, SAHA and belinostat,19 20 and benzamides such as MS275,21 have very limited penetration of the blood–brain barrier. Indeed, valproic and certain newer antiepileptic agents may exert their effects by inhibition of HDAC activity and differential effects on excitatory and inhibitory neuronal activities.22,,24 The ability to enter the brain or not under usual dosing is relevant to use of HDACi to treat central nervous system tumours or inflammatory conditions, whereas exclusion from the brain may be a useful consideration in avoiding potential (central nervous system) side effects or toxicity. However, for HDACi-associated modulation of inflammatory responses, the precise sites of action, cellular targets and mechanisms of action are not well understood, with a considerable number of in vitro and in vivo studies offering complex, incomplete and sometimes contradictory datasets. Resolution of this problem may take a long time, given the divergent approaches of scientists, ranging from studies of the effects of an HDACi on cells in vitro (table 1) to assessment in whole animals and complex disease models (table 2).

Table 1

Effects of HDACi on cells in vitro

Table 2

Effects of HDACi in vivo*

HDACi and inflammatory cells

Early studies showed that hydroxamate-based pan-HDACi compounds, such as TsA, SAHA and similar agents, have inhibitory effects on cytokine production by lipopolysaccharide (LPS)-treated monocytes cultured in vitro (table 1) as well as in mice injected with LPS, acting to dampen monocyte production of tumour necrosis factor (TNF) α, interleukin (IL)-1α and IL-1β.25 26 These early data of the effects of HDACi on monokine production and macrophage activation were subsequently confirmed27 and extended to additional areas. Thus, synovial macrophages and intact synovial tissue samples from patients with rheumatoid arthritis showed decreased cytokine production (IL-6, TNFα) and increased apoptosis when exposed to TsA in vitro.28 These data are consistent with the upregulation of HDAC expression by LPS and increased expression of proinflammatory genes, such as COX-2, though which combination of HDACs is upregulated appears to vary by cell type and, more broadly, by proinflammatory stimulus.29 30 In contrast with their effects on monocytes and dendritic cells (DC), the effects of HDACi on T cell responses are more nuanced than initially appeared, with evidence of HDACi inhibiting the proliferative and cytotoxic activity and cytokine production of CD3 monoclonal antibody-activated T cells,31 and impairing generation of signal 2 (costimulatory molecules) and signal 3 (activating cytokines) in antigen-pulsed antigen-presenting cells, leading to impaired proliferation and chemotaxis of Th1 but not Th2 cells.32 33

HDACi and disease models

The range of diseases beyond malignancies in which HDACi use has proved therapeutic, experimentally, is considerable, though various negative data may, of course, not be reported. Much work has been directed towards experimental studies of sepsis and autoimmunity, especially rodent models of arthritis and colitis (table 2). There are also some less expected areas in which use of HDACi has proved beneficial. For example, studies of the effects of HDACi use in rodents with genetically associated or salt-induced hypertension showed remarkable reduction in expression of proinflammatory cytokines, cardiovascular injury and stress responses.34 35 While HDACi administration also lowered blood pressure, the effects noted were not seen by use of a standard antihypertensive agent (hydralazine), and probably include suppression of inflammatory pathways that are activated by hypertension.34,,36 Additional areas include potential beneficial effects of HDACi treatment on the development of chronic inflammation and epithelial/mesenchymal transformation,37,,40 and also effects on the inflammatory process associated with the development of atherosclerosis.41 42

Mechanisms of action and caveats

At first, the use of compounds that promote acetylation of histone tails and increase accessibility to the DNA of various transcription factors may not appear likely to have useful anti-inflammatory effects. This assessment is also consistent with at least initial insights about one of the main proinflammatory pathways, involving NF-κB activation. Thus, much attention has been directed towards the effects of HDACi on the NF-κB pathway, and it seems that HDACi compounds can be either activators of proinflammatory genes or inhibitors based on their molecular target, the proinflammatory mediator used or the cell type.10 HDACi can downregulate NF-κB activation by inducing expression of IκBα, inhibiting its proteasomal degradation, and blocking the nuclear translocation of NF-κB and its DNA binding (figure 2), but other different effects can also occur. For example, p300/CREB-binding protein (CBP)-dependent acetylation at K310 of p65 prevents its association with IκBα and promotes DNA binding and transactivation.43 Deacetylation of this lysine is catalysed by HDAC3 (figure 2),43 44 such that selective HDAC3 inhibition might increase NF-κB activation, though this may occur in a gene-specific manner.45 The induction of multiple proinflammatory genes by p65 is also regulated by HDAC1 and HDAC2.46

Figure 2

Effects of HDACi on NF-κB and JAK/STAT pathways. NF-κB: Inflammatory stimuli activate T cell receptors (TCR), tumour necrosis factor receptors (TNFR), Toll-like receptors (TLR) or IL-1 receptors, which drive phosphorylation of NF-κB subunits. Phosphorylated IκBα is subject to proteasomal degradation, while phosphorylated p65 and p50 dissociate from IκBα and translocate to the nucleus. HDACi can inhibit proteasomal degradation of IκBα, and block the nuclear translocation of NF-κB and its DNA binding. In the nucleus, HAT enzymes (p300/CBP) acetylate NF-κB/p65 at K310, which promotes DNA binding and transcription of proinflammatory genes. HDAC3 deacetylates this lysine, and NF-κB is bound by newly synthesised IκBα, leading to nuclear export of NF-κB. HDACi use promotes IκBα synthesis. JAK/STAT: Inflammatory signals and cytokines such as IFNγ activate Janus kinase (JAK); JAK activates signal transducer and activator of transcription 1 (STAT1) by phosphorylation, leading to nuclear translocation of dimerised phospho-STAT1. In the nucleus, STAT1 binds to DNA and promotes transcription of IFNα-responsive genes. The acetylation of STAT1 by CBP destabilises this enhancesosome, leading to termination of IFN-γ-dependent STAT1 signalling. Acetylated but dephosphorylated STAT1 exits the nucleus. HDAC3 can deacetylate STAT1, which makes it available for reactivation. HDACi that inhibit HDAC3 promote STAT1 acetylation and thereby have potent anti-inflammatory effects. CBP, CREB-binding protein; HDACi, histone/protein deacetylase inhibitors. HAT, histone acetyltransferase; IL, interleukin; IFN, interferon.

The question thereby arises as to why selective targeting of an HDAC, such as HDAC3, promotes NF-κB activation and induction of many proinflammatory genes, while HDACi are under investigation as anti-inflammatory drugs? Despite all the empirical studies showing that HDACi do have important anti-inflammatory actions, this question remains unanswered, though clues are available. At least in some test systems, HDACi can stabilise IκBα expression and prevent its proteasomal degradation,47 48 and prevent the nuclear translocation and DNA binding of NF-κB (figure 2).49 Likewise, class I HDACs appear to be required for signal transducer and activator of transcription (STAT)-dependent transcriptional activation and proinflammatory gene expression.50,,53 Perhaps most critically, while tyrosine phosphorylation of STAT1 promotes dimerisation, nuclear translocation and activation of interferon γ-responsive genes, the acetylation of STAT1 by CBP destabilises this enhancesosome through recruitment of the tyrosine phosphatase TCP45, leading to termination of interferon γ-dependent STAT1 signalling.54 55 Acetylated but dephosphorylated STAT1 exits the nucleus and is deacetylated by HDAC3, and latent STAT1 is now available for reactivation. HDACi that inhibit HDAC3 promote STAT1 acetylation and thereby have potent anti-inflammatory effects. These data demonstrate that important exceptions exist to the concept that HDACs act as transcriptional repressors and histone acetyltransferases act as transcriptional activators, and help to explain the anti-inflammatory effects of HDACi in vivo (figure 2).

In addition to modifying chromatin accessibility or key signalling pathways, such as those involving NF-κB and Janus kinase (JAK)/STAT pathways, several further mechanisms may contribute to the anti-inflammatory actions of HDACi. First, HDACi may induce the apoptosis of cytokine-producing inflammatory cells,56 57 though there are relatively few data as yet for this mechanism of action using normal rather than transformed cells. Second, HDACi may disrupt the functional microtubule network in monocytes and thereby disrupt exocytotic release of cytokines from lysosomes.58 The significance of this mechanism is unclear given that the data were generated in vitro, with maximal effect on IL-1β release, lesser effect on TNFα and essentially no effect on IL-8 secretion upon LPS stimulation of cultured human monocytes. In addition, the effects of HDACi were reversed as doses of HDACi were increased, and the intracellular levels of cytokines were unaffected.58 Third, HDACi might affect the development and functions of cells with suppressive functions. There are a number of such cell types, beginning with the Foxp3+ T regulatory (Treg) cells, which are discussed in detail below, but also other populations of lymphocytes, including CD8+ suppressor T cells,59,,61 IL-10 producing Tr1 cells62,,64 and non-lymphoid cells, such as myeloid-derived suppressor cells.65 66 However, apart from effects on Foxp3+ Tregs, as yet, little is known of the effects of HDACi on these various suppressor cell populations. Also unexplored are the possible effects of HDACi on micro-RNA production and the stability of mRNA encoding inflammatory mediators.

In addition to acknowledging the uncertainties as to how HDACi exert their anti-inflammatory responses, it should be noted that important caveats exist as to their use. For example, at least in vitro, hydroxamates such as TsA and SAHA can potentiate microglial production of proinflammatory mediators, in association with enhanced NF-κB activation,67 though use of TsA and comparable hydroxamates decreased injury in vivo in murine models of experimental allergic encephalomyelitis68 and neuroinflammation.69 Theoretically, such agents may also exacerbate acute and chronic respiratory diseases,70,,73 and increased histone acetylation and proinflammatory gene expression are reported in asthma.74 75 Moreover, deacetylation of the glucocorticoid receptor by HDAC2 appears necessary for optimal responses to steroid treatment in steroid-resistant asthma.76 These outcomes sem to be related to gene induction by acetylation and also by removal of the inhibitory effect of one or more HDACs on expression of proinflammatory genes. A notable example of the latter is the finding that both high glucose and HDACi decrease HDAC1 binding to the TNFα promoter and increase TNFα expression by monocytes from patient with type 1 diabetes.77 Thus, there are largely in vitro and/or descriptive data suggesting the need for caution in use of HDACi as anti-inflammatory agents for disease of specific organs or throughout the vasculature,78 but in each case in vivo experimental data often support this application, illustrating how much needs to be learnt before this complex and multifaceted puzzle can be resolved.

HDACi and FOXP3+ Tregs

The clinical use of many pan-HDACi is associated with a common adverse effect profile of cardiac QT prolongation, nausea, diarrhoea, vomiting, hypokalaemia, loss of appetite and thrombocytopenia, plus in many cases, profound and debilitating fatigue. Likewise, their ability to induce cytotoxicity is considered a key and highly desirable action in the context of malignancies, which often overexpress HDAC1 and HDAC2, but the toxicity profile and cytotoxic effects render these agents far less suitable for non-oncological applications. To that end, various groups are seeking to avoid the class-associated side effects of pan-HDACi by trying to design isoform-selective HDACi for use in oncology and inflammation.

Our focus on selective HDACi arose from our findings in testing several HDACi compounds for their effects in murine models of colitis, including dextran sodium sulphate-induced colitis and the T cell-dependent CD45RBhi adoptive transfer model.79 80 Two pan-HDACi compounds, TsA and SAHA, but not MS275, a potent and long-acting HDAC class I-specific inhibitor, blocked development of colitis as shown by prevention of weight loss and associated blood in the stool, diarrhoea and histological injury. Likewise, in T cell-dependent adoptive transfer models, both pan-HDACi but not MS275 were effective in preventing the development of colitis, and in promoting the resolution of established colitis. The beneficial effects of pan-HDACi were dependent upon the presence of Foxp3+Treg cells, since Treg depletion or use of Scurfy mice with a mutation in Foxp3 abrogated any therapeutic benefit of HDACi administration.80 Foxp3+Tregs play a key part in limiting autoimmunity and maintaining peripheral tolerance, and mutations of Foxp3 lead to lethal autoimmunity in humans and mice.81,,85 In wild-type (WT) mice, pan-HDACi but not MS275 use decreased mucosal inflammatory cytokine production, and increased Foxp3 and anti-inflammatory cytokine expression, and enhanced Treg suppressive function.79 80

Further in vitro analysis86 demonstrated that multiple pan-HDACi hydroxamates such as TsA, SAHA, M344 and scriptaid were effective in low nanomolar levels at enhancing murine Treg function, as well as the suppressive functions of rhesus macaque87 and human88 Treg cells. Additional pan-HDACi, such as the short-chain fatty acids, phenylbutyrate and valproic acid, also enhanced murine Treg function, but were only active in the micromolar and millimolar ranges, respectively. Our findings for the TsA-induced in vivo expansion of Foxp3+ Treg numbers and function were confirmed by other groups,89,,92 as was the induction of Foxp3 using other HDACi, such as SAHA.93 In contrast to our data using pan-HDACi, we found that class I-specific HDACi, such as the benzamides, MS275 and MC1293, and the quinolinol, NSC3852, lacked any effect on Treg functions in vitro when used at micromolar or higher levels.80 86 Hence, at least when using standard therapeutic dosages, only agents that blocked both class I and class II HDACs were effective at enhancing Treg function, and since class I-selective HDACi compounds were ineffective in the same assays, our data point to a key role for class II HDAC in control of Treg functions.

Targeting class IIa HDACs

Compared with the extensive literature on pan-HDACi, the identification of HDAC class- or subclass- or isoform-selective inhibitors is in its infancy. Class I-selective HDACi include MS27594 and MC129395 noted above, as well as 4-phenylimidazole,96 and compounds selective for HDAC197 are reported. Class II-selective HDACi include MC1568 and MC1575; these were originally reported as class IIa-selective but are now known to also inhibit the class IIb HDAC, HDAC6.98 99 HDAC isoform-specific inhibitors include agents with a high selectivity for HDAC4,100 HDAC6,101,,107 or HDAC8.108 109 This section will consider aspects of class IIa HDAC biology and therapeutic targeting.

Class II HDAC lack potent catalytic activity when assayed using conventional acetyl-lysine peptide substrates,110 in large part because of the presence of a tyrosine in the active site, instead of a histidine as occurs in class I and class IIb HDACs; mutation of this tyrosine to histidine improves HDAC activity against conventional acetyl-lysine substrates 1000-fold.111 The development of alternate non-acetyl-lysine (trifluoroacetate) substrates has allowed identification of significant catalytic activity of class IIa HDACs,112 but the physiological relevance of this activity remains unclear, especially since class I and IIb HDACs are inactive against these alternative substrates.110 113 A developing view is that class IIa HDACs serve as recognition units or receptors for acetylated lysines113 and function by recruiting class I HDACs, especially HDAC3,114 115 thereby providing deacetylating activity. This recruitment involves residues within a zinc-binding subdomain conserved only in class IIa HDACs, such that small molecules that bound to the active site of class IIa HDACs can disrupt interaction with HDAC3/N-CoR repressor complexes and block the associated catalytic activity provided by HDAC3 to class IIa HDACs via protein–protein interactions.114 115 However, additional recruitment of HDAC3 via N-terminal binding of COOH-terminal-binding protein (CtBP) may also need to be targeted to effectively block HDAC3/class IIa HDAC interactions.116

Our interest in class IIa HDACs arose from our finding of prominent expression of HDAC9 in murine79 and human88 Foxp3+Tregs, and that gene targeting or siRNA knockdown of HDAC9 enhanced Treg suppressive function in vitro and in vivo (figure 3).79 80 Likewise, gene targeting of HDAC7 increases Treg suppressive functions in vitro and in vivo (figure 3).117 118 Microarray studies indicated that the effects of HDAC7 and HDAC9 on Treg gene expression were distinct, such that targeting of both might have therapeutic potential, but how to undertake this is not clear. We are currently undertaking an analysis of the effects of HDAC3 targeting, since many of the actions of class IIa HDACs may be attributable to the deacetylase activity of HDAC3.114 115 Nevertheless, these class IIa HDAC proteins are known to regulate gene expression through protein/protein interactions.119 120 Moreover, even in the case of HDAC9, a rather poorly understood and little studied HDAC, genes regulated by HDAC9 as a result of its apparently weak catalytic activity are beginning to be identified.121 122 Hence, the development of small molecules that inhibit the low catalytic activity of one or more class IIa HDACs, or disrupt their protein/protein interactions, may eventually show therapeutic potential as new types of anti-inflammatory HDACi.16

Figure 3

Effects of HDAC targeting in Tregs. Diagram summarising the known mechanisms by which HDAC6, HDAC7 and HDAC9-/- targeting increases Treg suppressive function in vitro and in vivo. HDAC, histone/protein deacetylase; Treg, T regulatory.

Targeting class IIb HDACs

Localised primarily to the cytoplasm, HDAC6 regulates the acetylation of multiple proteins, such as α-tubulin and heat shock protein 90 (HSP90), but also has deacetylase-independent functions.123,,127 Unique in the field of HDACi, multiple HDAC6 isoform-selective HDACi (HDAC6i) are reported.101,,107 These considerations led us to explore the effects of HDAC6 targeting on Tregs and, by extension, whether isoform-selective HDAC6i might be useful as anti-inflammatory agents.128 We found that HDAC6 was expressed at several-fold higher levels in Tregs than in conventional T cells, and HDAC6 knockout mice thereby provided a ‘gold standard’ as to how effective pharmacological inhibitors of HDAC6 might be expected to be in modulating immune events. HDAC6-/- mice are known to be immunocompetent and not prone to tumourigenesis or chronic infections.129 However, their Tregs were more suppressive in vitro and in vivo than WT Tregs (figure 3). Although HDAC6-/- Tregs express more Foxp3, CTLA4 and IL-10 than their WT counterparts, the basis for this increased suppressive capability may be multifaceted. HDAC6 gene targeting would probably disrupt both the deacetylase-dependent and -independent functions of normal HDAC6. The latter include roles for HDAC6 in regulation of cell migration and proteasomal degradation.

Evidence of the effects of HDAC6 targeting on deacetylase-dependent functions was readily apparent in our studies, including hyperacetylation of HSP90 and the upregulation of many HSF1-regulated genes in HDAC6-/- Tregs, including multiple HSP family members. While many additional genes of importance to Treg biology, but without known regulation via HSF1, were also differentially expressed in HDAC6-/- Tregs, effects on the HSF1/HSP pathway are likely of major importance both mechanistically and therapeutically. We have recently shown that HSP70 forms a complex with Foxp3 in Tregs, that upregulation of HSP70 promotes Treg survival and suppressive functions under conditions of cell stress, and that inhibition of HSP70 impairs Treg survival and suppressive functions.80 The current data point to a major role for intracellular heat shock responses in control of Treg functions.

We found in colitis and transplant models that the presence or absence of HDAC6 just within Tregs is a powerful determinant of Treg-dependent resolution of colitis and resistance to allograft rejection.128 These data underline the importance of HDAC6 as a therapeutic target for modulation of Treg responses. Analysis of the effects of HSP90i in vitro and in vivo in our studies indicated that at least for the models under consideration, targeting of HDAC6 or HSP90 had broadly comparable effects and did not show obvious additional benefits when used together. Such combination might allow for lower doses of each inhibitor to be used, but the broad message from our work so far is that the benefits of targeting the HSF1/HSP pathway appear to be achieved by pharmacological modulation of HDAC6 or HSP90. Some 14 HSP90i compounds, including 17-allylamino, 17-demethoxygeldanamycin, are currently being evaluated in phase 1 and phase 2 clinical trials; while data are preliminary, toxicity was rarely seen.130 131 Clinical development of HDAC6i is less developed, but HDAC6 targeting is being considered as a treatment for neurodegenerative conditions.107

Our finding that selective targeting of an individual HDAC isoform can provide comparable effects on Tregs, and associated suppression of T cell-dependent immune responses, to that seen using broadly acting pan-HDACi provides a powerful rationale for the ongoing evaluation of HDAC6i in the regulation of inflammation. Ultimately, selective HDAC6i may provide an alternate, pharmacological approach to treatments dependent upon Treg expansion and adoptive transfer for the management of autoimmunity and transplant rejection.

Targeting class IV HDAC

In the first evidence for a physiological function for the sole class IV HDAC, HDAC11, data from gene targeting and siRNA approaches showed that HDAC11 expression suppressed macrophage production of IL-10.132 HDAC11 has antiproliferative effects133 134 and is upregulated in at least some cell types, such as pancreatic beta cells, by cytokine simulation.135 As for class IIa HDACs, no specific inhibitors of HDAC11 have been reported. However, HDAC11 may be present in complexes that also contain HDAC6,136 137 such that studies of the effects of HDAC6i on the biology of HDAC11-/- mice, including in models of inflammation, may be informative.


HDACi compounds act in cancer models by inhibiting the cell cycle, inducing apoptosis and limiting angiogenesis. While HDACi probably exhibit the same effects in models of inflammation, the relative importance of these actions is likely to be markedly different. HDACi exhibit anti-inflammatory effects in a remarkable variety of models and contexts, although their effects on macrophages and DCare such that Th1-dependent responses are more commonly suppressed than Th2-dependent responses, at least in models reported to date. There are also new mechanisms that involve further cell types than the commonly studied antigen-presenting cells and T cells. These include clinically important effects on the acetylation of Foxp3 and potentiation of Foxp3+Treg-dependent immune suppression. Ongoing studies to further dissect and target individual HDAC isoenzymes are underway and may have important advantages over the predominant one-size-fits-all strategy of using pan-HDACi. In particular, targeting of HDAC6 using selective HDAC6i has considerable therapeutic potential in inflammation.


  • Funding National Institutes of Health. Supported in part by a research grant from the National Institutes of Health to WWH (P01AI073489).

  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.


Free sample This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of ARD.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.