2-(Difluoromethyl)-1,3,4-oxadiazoles: The Future of Selective Histone Deacetylase 6 Modulation?

Abstract

Histone deacetylase 6 (HDAC6) is an important target for the treatment of oncological and non-oncological diseases. Established HDAC6 inhibitors feature a hydroxamic acid as a zinc-binding group (ZBG) and thus possess mutagenic and genotoxic potential. Recently, the 2-(difluoromethyl)-1,3,4-oxadiazole (DFMO) group emerged as a novel ZBG. In this Viewpoint, we summarize the discovery of the mode of action of DFMOs. Additionally, we discuss opportunities and challenges in the journey toward the clinical development of DFMO-based drugs for the treatment of HDAC6-driven diseases.

Histone deacetylases (HDACs) play a pivotal role as epigenetic regulators, thereby holding significant promise as therapeutic targets for a range of conditions, including cancer, inflammation, and neurodegenerative diseases.¹ The HDAC family is divided into four classes with a total of 18 isoforms.¹ Classes I, II, and IV are Zn²⁺-dependent, whereas class III enzymes, also known as sirtuins, are NAD⁺-dependent.² Class I consists of HDAC1–3 and HDAC8, which are mostly located in the nucleus and mainly catalyze the deacylation of lysine side chains of histones.^3,4 Class II is further subdivided into class IIa, consisting of HDAC4, 5, 7, and 9, and class IIb, containing HDAC6 and 10.⁵ While class IIa HDACs shuttle between the nucleus and the cytoplasm, their counterparts in class IIb are primarily located in the cytoplasm and deacetylate lysine side chains of non-histone proteins or, in the case of HDAC10, polyamines such as spermidine as their primary substrates.^1,6 Further, the only member of class IV, HDAC11, is mainly located in the nucleus.²

HDAC6 stands out among HDAC isoforms due to its distinct structural features, localization, and substrate range.⁵ Notably, HDAC6 features two independently active catalytic domains (CD1 and CD2) and a zinc finger serving as an ubiquitin-binding domain.^1,7 Initially identified as a tubulin deacetylase, HDAC6 also catalyzes the deacetylation of various other proteins.⁸ Moreover, it plays a pivotal role in modulating cortactin,⁹ the Alzheimer-related tau protein,¹⁰ and the chaperone HSP90.^1,11 Consequently, HDAC6 is crucial for various cellular processes (see Figure 1 for representative examples), including cell motility, proliferation, apoptosis, and the aggresomal pathway, thereby making it a valuable target for drug development.¹

Figure 1.

HDAC6 interactions in specific cellular pathways. (A) HDAC6 inhibition induces α-tubulin acetylation and microtubule stabilization and diminishes cancer cell migration. (B) HDAC6 inhibition leads to HSP90 hyperacetylation, accumulation of misfolded protein, and degradation of HSP90 client proteins. (Figure created with BioRender.com.)

Commonly, HDAC inhibitors (HDACi) consist of a zinc-binding group (ZBG) binding to the Zn²⁺ ion in the catalytic site, a cap group for protein surface interactions, and a linker connecting these two moieties.¹² The non-selective HDACi vorinostat, belinostat, panobinostat, and romidepsin (see Figure 2A), previously approved by the U.S. Food and Drug Administration (FDA), lack selectivity across HDAC isoforms.^1,13 This may lead to severe side effects and off-target interactions.¹³ In recent years, selective HDAC6 inhibition was primarily achieved by utilizing HDACi with a bulky cap group, a phenyl or benzyl linker, and a hydroxamic acid as ZBG (Figure 2B), which provided T-shaped inhibitors capable of engaging the unique and relatively wide L1 loop pocket exclusively present in HDAC6.¹² However, the potential of hydroxamic acids for generating mutagenic and genotoxic metabolites, through either Lossen rearrangement or hydroxylamine release, underscores the urgent need to identify novel ZBGs.¹⁴ Recently, the 2-(difluoromethyl)-1,3,4-oxadiazole (DFMO) group has emerged as a promising alternative to hydroxamic acids; for selected DFMO-based HDAC6 inhibitors see Figure 2C.¹⁵⁻¹⁹ These novel inhibitors are gaining their high isoform specificity by serving as substrate analogs and thus mechanism-based inhibitors of HDAC6. In this Viewpoint, we summarize the discovery of the unique mode of action of DFMOs as potent and selective HDAC6 inhibitors. Furthermore, we discuss our views on opportunities and challenges concerning the clinical development of DFMO-based drugs for the treatment of HDAC6-driven diseases.

Figure 2.

(A) FDA-approved HDAC inhibitors. (B) Selected hydroxamate-based HDAC6 inhibitors. (C) Selected DFMO-based selective HDAC6 inhibitors: BK-1,¹⁵ SE-7552,¹⁶ ITF5924,¹⁷ Cmpd7,¹⁸ Cmpd9,¹⁹ and T-518.²⁰

History of DFMO-Based Selective HDAC6 Inhibitors and Degraders

The DFMO group was first introduced as a ZBG for selective HDAC6 inhibition in a 2017 patent by Chong Kun Dang Pharmaceutical Corp.²¹ Despite its frequent appearance in patents, this ZBG is relatively underrepresented in research manuscripts. However, in 2021, Onishi et al.²⁰ disclosed that the DFMO-based selective HDAC6 inhibitor T-518 (Figure 2C) demonstrated therapeutic potential to treat Alzheimer’s disease and tauopathy in mice after oral administration. The high HDAC6 selectivity was confirmed at both the biochemical and cellular levels.²⁰ Furthermore, T-518 showed an encouraging pharmacokinetic profile and favorable brain penetration.²⁰ Shortly later, the hydroxamate tubastatin A (Figure 2B) and the DFMO derivative SE-7552 (Figure 2C), a compound first mentioned in a conference abstract as a non-hydroxamate HDAC6 inhibitor capable of blocking multiple myeloma growth in vivo,²² were used by Cone and co-workers as selective HDAC6 inhibitors to overcome leptin resistance in obesity.¹⁶ Notably, they could show that SE-7552 acts as an anti-obesity agent in diet-induced obese mice.¹⁶ In the next step, our group successfully incorporated the DFMO warhead into proteolysis-targeting chimeras (PROTACs) for the selective degradation of HDAC6.²³ However, the mechanism by which DFMOs inhibit or degrade HDAC6 remained enigmatic.

In a 2022 conference abstract, the Christianson group, in collaboration with us, disclosed that the DFMO derivative BK-1 (Figure 2C) underwent an enzyme-catalyzed ring-opening reaction.²⁴ This process led to the formation of an acylhydrazide, which was subsequently co-crystallized in an extended conformation within the active site of HDAC6.²⁴ In 2023, Steinkühler and colleagues elucidated the HDAC6 complex structure with a hydrazide inhibitor derived from the twofold hydrolysis of the DFMO inhibitor ITF5924 (Figure 2C).¹⁷ The authors speculated that the crystallized hydrazide, though identified, may not be responsible for the observed profound HDAC6 inhibition, suggesting that the hydrazide might not represent the primary active species.¹⁷ Instead, the authors postulated the presence of a high-affinity intermediate, forming a durable tight-binding enzyme–inhibitor complex.¹⁷ This intermediate may appear as a closed hydrated form or a protonated acylhydrazide; both options were suggested as potential active species.¹⁷ Almost in parallel, based on crystallographic and mechanistic experiments, the Christianson and Hansen groups disclosed the full experimental details of their initial conference contribution,²⁴ confirming that the DFMO warhead undergoes an enzyme-catalyzed ring-opening reaction, resulting in a deprotonated difluoroacetylhydrazide as active species (Figure 3).¹⁵ The strong anionic zinc coordination of the deprotonated difluoroacetylhydrazide and the binding of the difluoromethyl moiety in the P571 pocket of the CD2 of HDAC6 finally results in an essentially irreversible inhibition of the enzyme.¹⁵ The tight-binding properties of the active species were confirmed by jump-dilution and dialysis experiments.¹⁵ Overall, there is now clear evidence that DFMOs act as mechanism-based, slow- and tight-binding HDAC6 inhibitors.¹⁵

Figure 3.

Schematic representation of the enzyme-catalyzed ring-opening reaction of inhibitor BK-1 by HDAC6 (based on PDB ID 8GD4).

Also in 2023, Barinka and co-workers performed a systematic comparison of the DFMO Cmpd7 (Figure 2C) and its corresponding hydroxamic acid analog, thereby clearly supporting the observation that DFMOs are new ZBGs with unparalleled selectivity for HDAC6 over all other HDAC isoforms.¹⁸ Additionally, a comprehensive mechanistic analysis of the hydrolysis of Cmpd7 provided additional evidence that DFMOs can undergo an enzyme-catalyzed ring-opening reaction to a difluoroacetylhydrazide as well as a significantly slower second hydrolytic step to the corresponding unsubstituted hydrazide.²⁵ Another thorough biochemical evaluation of ITF5924 further confirmed the slow- and tight-binding properties of DFMOs.²⁶ Similarly, Ripa et al.¹⁹ investigated the bioactivation of DFMOs and confirmed once more that this class of compound acts as a substrate analog of acetyl-lysine. Their crystallization attempts of the DFMO Cmpd9 (Figure 2C) confirmed that the DFMO was degraded to the corresponding hydrazide.¹⁹ Furthermore, the authors showed that DFMOs are stable at pH = 7 but chemically unstable at acidic and basic pH values.¹⁹ In addition, Cmpd9 showed a promising in vitro safety profile with no activity in a panel of cardiovascular ion channels and was negative for in vitro micronuclei, thereby indicating that this DFMO derivative possesses neither cardiovascular toxicity nor genotoxicity.¹⁹ Importantly, this study also disclosed a high oral bioavailability and low in vivo clearance of DFMO-based selective HDAC6 inhibitors.¹⁹

Conclusion and Future Outlook

DFMO-based HDAC6 inhibitors did not appear in the scientific literature until recently. However, in the past few years, several publications on the HDAC6 selectivity and mode of action appeared. While the first papers did not address the mode of action of DFMOs, recent work from both academic groups and industry has confirmed that DFMOs act as substrate analogs and thus are mechanism-based inhibitors of HDAC6. Currently, there are several HDAC6-selective inhibitors with undisclosed structures in clinical trials. This raises the question of whether DFMO-based HDAC6 inhibitors have already progressed to the clinical stage. However, there are several challenges ahead to obtain regulatory approval. The chemical stability of DFMO derivatives is the most obvious challenge during their clinical development.¹⁹ The limited stability at acidic and basic pH values arises from the electrophilicity of the oxadiazole ring. While it might be possible to optimize the stability further by fine-tuning the substitution in the 2- and 5-positions of the 1,3,4-oxadiazole ring, it is rather unlikely to mitigate the stability issues completely, because the electrophilicity is required for the mode of action relying on being a substrate analog capable of undergoing the essential ring-opening reaction. Nevertheless, DFMOs are stable at pH = 7, and suitable formulations have been developed for various acid-labile drugs, including blockbusters such as proton pump inhibitors.

Besides that, the mode of action of DFMOs might not be perfectly suitable for the incorporation of this warhead into PROTACs. One major advantage of PROTACs is their catalytic mode of action. In the case of DFMO-based HDAC6 PROTACs, the oxadiazole ring will undergo ring-opening during the first degradation cycle, and the degrader will presumably be released as a hydrazide featuring a reduced ability to engage HDAC6. However, Steinkühler and co-workers reported that the corresponding hydrazide obtained from ring-opening of ITF5924 followed by a second hydrolytic step is capable of inhibiting HDAC6 with nanomolar potency.¹⁷ Three co-crystal structures of degraded DFMOs (PDB IDs 8CJ7, 8BJK, and 8A8Z) provide further evidence that the resulting final hydrolysis products, the respective hydrazides, can reengage HDAC6, albeit presumably with somewhat lower degradation efficiency.^17,19,25 Nevertheless, the first DFMO-based PROTACs reduced HDAC6 levels with DC₅₀ values in the low triple-digit nanomolar concentration range, thus suggesting that DFMOs are suitable for the development of non-hydroxamate HDAC6 degraders.²³

Despite the challenges ahead, there are compelling reasons to be optimistic that DFMOs can succeed in clinical trials. These include favorable pharmacokinetic profiles and efficient brain penetration, along with high oral bioavailability and low in vivo clearance. Notably, DFMOs exhibit no cardiovascular toxicity or genotoxicity, providing additional evidence for their potential safety and efficacy in clinical applications.¹⁹ Moreover, these inhibitors showcase an outstanding selectivity profile within the HDAC family and beyond, positioning DFMOs as highly promising candidates for the treatment of cancer and other HDAC6-related diseases.^18,20 Additionally, the proven in vivo efficacy against diverse conditions such as multiple myeloma,²² obesity,¹⁶ and tauopathies²⁰ underscores the significant potential of DFMOs. To conclude, the future could be bright for DFMO-based HDAC6 modulators.

Glossary

Abbreviations

CD

catalytic domain

DFMO(s)

2-(difluoromethyl)-1,3,4-oxadiazole(s)

FDA

U.S. Food and Drug Administration

HDAC(s)

histone deacetylase(s)

HSP90

heat shock protein 90

NAD⁺

nicotinamide adenine dinucleotide

PROTAC(s)

proteolysis targeting chimera(s)

ZBG(s)

zinc-binding group(s)

The authors declare no competing financial interest.