Ferrocene Derivatives as Histone Deacetylase Inhibitors: Synthesis and Biological Evaluation

Abstract

Ferrocene is an organometallic compound that has attracted considerable scientific interest due to its unique properties, including low toxicity, excellent stability in aqueous and aerobic media, and high lipophilicity, which enhances membrane permeability. The ferrocene moiety has been effectively used as a bioisostere of phenyl rings and heteroaromatic groups in the structures of approved tyrosine kinase inhibitors and histone deacetylase inhibitors (HDACis). HDACis exert their cytotoxic effects by blocking cyclin/CDK complexes, causing cell cycle arrest, inducing apoptosis, inhibiting angiogenesis, and through non-histone-directed mechanisms. This mini-review summarizes the synthesis and biological evaluation of small libraries of compounds in which a ferrocenyl moiety is incorporated into the structure of suberoylanilide hydroxamic acid (SAHA) and a number of analogues. The influence of the organometallic function on the antiproliferative effect is investigated. Both docking analysis and in vitro studies confirm that the ferrocenyl-modified HDACis exhibit potent cytotoxicity and strong inhibitory activity against the various enzyme isoforms.

1. Introduction

The enzymes histone acetylase and histone deacetylase are crucial for gene expression [1]. Both enzymes control chromatin structure and DNA packaging in the cell nucleus, exerting their effects during nucleosome formation [2]. Histone acetylase catalyzes the acetylation of lysine residues in histone side chains, neutralizing their positive charge. This neutralization weakens the ionic bonds between histones and the negatively charged phosphate groups in the DNA backbone. As a result, the chromatin structure becomes less compact, facilitating access for transcription factors to bind to their specific DNA sequences [3]. Histone deacetylase, on the other hand, removes these acetyl groups, leading to increased chromatin condensation. The dysregulation and overactivity of histone deacetylase are associated with gene silencing [4].

The cytotoxic effect of histone deacetylase inhibitors (HDACis) stems from multiple cellular mechanisms. A key mechanism is the induction of apoptosis, or programmed cell death. These inhibitors affect genes involved in both the intrinsic and extrinsic apoptotic pathways, ultimately leading to cell death [5,6,7]. The antiproliferative activity of this class of compounds is also linked to their ability to block the formation of cyclin/CDK complexes, inhibit angiogenesis (thereby preventing tumour vascularization), induce autophagy, and modulate the immune response against tumour cells [8,9,10,11]. Vorinostat, also known as suberoylanilide hydroxamic acid (SAHA), is the most typical HDACi and the first one approved for the treatment of cutaneous T-cell lymphoma (CTCL) [12].

Ferrocene (Figure 1) is a unique organometallic compound, characterized with a number of advantages that make its framework attractive for application in medicinal chemistry [13]. Ferrocenes have low toxicity, excellent stability in aqueous and aerobic media, and high lipophilicity, which enhances membrane permeability [14,15]. Ferrocene derivatives have demonstrated antimalarial, antimicrobial, and anticancer activity [16,17,18,19]. The ferrocene moiety has also been effectively used as a bioisostere for phenyl and heteroaromatic groups in the structures of tyrosine kinase inhibitors [20,21,22]. The potential of ferrocene as an anticancer agent was first recognized in the late 1970s, when Fiorina et al. demonstrated the antitumor activity of ferrocene derivatives bearing amino or amide groups against P-388 leukemia. Their findings unequivocally showed that incorporating a ferrocene fragment into a molecule enhances antitumor activity [23]. A significant development of the concept was made by Jaouen and coworkers who modified the structure of tamoxifen by replacing a phenyl ring with ferrocene. This led to the discovery of ferrocifens, compounds with a broader spectrum of anticancer activities than the parent drug [24].

The incorporation of ferrocene as a bioisostere has also proved to be a successful strategy in the design of novel molecules with histone deacetylase inhibitory activity. This mini-review aims to provide a concise overview of recent developments in the synthesis and biological evaluation of the ferrocene HDACis, compounds with structures largely based on SAHA, incorporating a ferrocenyl moiety.

2. Ferrocene Derivatives as HDACis

In 2011, Spencer and coworkers developed the first organometallic analogues of SAHA (Figure 2). They used a ferrocenyl group as phenyl bioisostere and studied the effect of the organometallic moiety on the antiproliferative activity [25]. Jay amine hydroxamic acid (JAHA) exhibited a nanomolar inhibition of HDACs, as well as selectivity and cytotoxicity against breast cancer cell lines. Subsequently, a small library of JAHA analogues was synthesized, and the initial results of the structure–activity relationship (SAR) studies were presented.

The synthesis of JAHA was carried out in two steps. The microwave-mediated reaction of ferrocenylamine 7 with methyl 8-chloro-8-oxooctanoate in the presence of triethylamine as the base afforded the ester 8 in good yields. A subsequent reaction with hydroxylamine, prepared in situ from its hydrochloric salt, gave the desired hydroxamic acid (Figure 3). JAHA analogues 3–6 were synthesized for the purpose of SAR studies using analogous reactions.

Molecular docking investigations revealed that JAHA binds to the catalytic centre of HDAC8 in a similar manner to SAHA. The hydroxamic acid coordinates with the zinc ion at the bottom of the enzyme’s active centre. The ferrocene functionality is situated at the entrance of the pocket, which has sufficient space to accommodate the bulk metallocene moiety. Additionally, a hydrogen bond is observed between the amide moiety of the ferrocene-containing analogue and the enzyme. The results of the docking assay were confirmed by the finding that JAHA had an extremely similar activity profile to the reference SAHA in terms of HDAC class I and II inhibition.

Compound 3 (homo-JAHA) was synthesized to elucidate the role of the methylene spacer in the structure of JAHA. Homo-JAHA exhibited a similar inhibitory activity towards class I HDACs 1–3, and a higher activity towards HDAC8, but a lower activity towards class II HDAC6. The aryl-linker JAHA analogues 4–6 were synthesized to evaluate the effect of the organometallic group directly attached to SAHA. The docking analysis of compound 4 showed a better fit to the hydrophobic pocket of HDAC8, suggesting higher binding affinity to the enzyme in the background of the available van der Waals interactions between the ferrocene group and the amino acid residues in the pocket. This was confirmed by the potency against HDAC8, with compound 4 showing the lowest value of IC₅₀ = 2 nM. Compound 5 significantly outperforms the other ferrocenes as well as SAHA against HDACs 1, 2 and, in particular, against class IIb, HDAC6 (IC₅₀ = 0.09 nM). Compound 6, an analogue of 5 with a shorter linker, showed poor enzyme inhibition, highlighting the importance of chain length on HDAC inhibition. The compounds were tested for their antiproliferative activity against the MCF-7 breast cancer cell line and were found to be less cytotoxic than SAHA, which was explained by reduced cellular penetration due to the presence of the ferrocene group. Additionally, the K562 cell line was cultured with the most promising compounds 4 and 5 and histone, and tubulin acetylation was quantified by flow cytometry. The activity of the novel compounds was comparable to that of SAHA. These initial results demonstrated that modification of the aromatic moiety of SAHA using the bulky planar ferrocene is a promising strategy for the synthesis of compounds with good inhibitory activity against HDACs [25].

HDAC inhibitors exert their cytotoxic effect primarily by restoring the expression of “silenced” genes, inducing cell death and cell cycle arrest in cancer cells. Nevertheless, they also exert their effect through non-histone-mediated mechanisms, such as interfering with the function of the p53 protein [26]. The in vitro antiproliferative activity of JAHA 2 and homo-JAHA 3 was evaluated in MDA-MB-231 cells [27]. The cell line was selected for investigation due to its characteristics as a triple-negative breast carcinoma. This refers to the lack of expression of the HER2 protein, the estrogen receptor (ER), and the progesterone receptor (PR), as well as an extremely high level of invasiveness and metastatic potential. Furthermore, the cell line exhibits a mutated form of the p53 protein, associated with an unfavourable prognosis for patients [28]. The results indicated that JAHA exhibited prominent antiproliferative activity against MDA-MB-231, with an IC₅₀ = 8.45 μM at 72 h of treatment. In contrast, homo-JAHA induced a reduction in cell number of only 63%. This was an indication that the additional methylene spacer in homo-JAHA had a negative effect on activity. The cytotoxicity of JAHA against MDA-MB-231 was found to be based on multiple mechanisms of action, including the production of reactive oxygen species (ROS) followed by mitochondrial membrane dissipation and autophagy inhibition, thus depriving tumour cells of a protective mechanism to cope with metabolic stress [27,29].

Further investigation of the cytotoxic effect of JAHA, as opposed to SAHA, against MDA-MB-231 revealed alternative mechanisms of action, namely the inhibition of DNA methyltransferases DNMT1 and DNMT3b [30]. Both enzymes are responsible for the silencing of tumour suppressor genes [31]. The authors proposed that the DNA demethylation that occurs following JAHA treatment results in the rearrangement of the molecular landscape of transcriptional regulation and the restoration of gene expression [30]. In addition, the cytotoxic effect of JAHA compared to homo-JAHA and SAHA on MDA-MB-231 breast cancer cells was studied in detail using a variety of in vitro assays. The remarkable activity of JAHA led to investigations of its mechanism of action, focusing in particular on cell cycle perturbation and the onset of apoptosis and/or autophagy, coupled with ROS overproduction and/or mitochondrial membrane dissipation [32].

Gasser and coworkers further developed the JAHA analogues into enzyme inhibitors with selective photorelease [33]. They blocked the hydroxamic acid moiety of Fc-SAHA with a photolabile protecting group (PLPG). The objective was to prevent the occurrence of toxicity in healthy cells and tissues. This concept is based on the use of light as an activator of the cytotoxic compound without the need of oxygen, and is often referred to as photoactivated chemotherapy [34]. It allows targeted delivery, which makes it a promising approach for the treatment of solid tumours where hypoxia is common [33,35].

Fc-SAHA 5 was selected as one of the first organometallic and, more importantly, light-stable HDAC inhibitors. Its hydroxamic acid group was photocaged by nucleophilic substitution with 1-(bromomethyl)-4,5-dimethoxy-2-nitrobenzene as a photolabile protecting group to give p-Fc-SAHA 9 (Figure 4). Photorelease kinetics and by-product formation were then evaluated. It was shown that UV-A irradiation (350 nm) leads to photoactivation and the successful release of Fc-SAHA. As expected, p-Fc-SAHA showed a significantly reduced inhibitory capacity against the tested HDAC isoforms (IC₅₀ = 3.45 for HDAC1, 209 for HDAC2, 10 for HDAC6) compared to Fc-SAHA (IC₅₀ = 0.13 for HDAC1, 0.33 for HDAC2, and 0.03 for HDAC6). The inhibitory activity of the sample was fully restored after UV-A irradiation, with IC₅₀ values of 0.11, 0.23, and 0.04 for HDAC1, HDAC2, and HDAC6, respectively. The residual activity of p-Fc-SAHA was found at higher concentrations due to insufficient steric hindrance of the active site of the enzyme by the photolabile protecting group. An obstacle to in vitro testing of the developed system on cell lines was the poor water solubility of p-Fc-SAHA [33].

In 2012, Spencer et al. reported conformationally restricted ferrocene-based HDACis [36]. The authors sought to modify isoform selectivity, improve pharmacokinetic properties, and reduce enzyme binding entropy by synthesizing triazole-containing click analogues of JAHA. The synthesis was conducted in two directions (Figure 5). The first approach was to introduce a triazole function in close proximity to the ferrocene “cap” of JAHA. This was achieved by the cycloaddition of ethynylferrocene 10 with two azides of different chain lengths, resulting in the formation of click products 11a,b. The latter were then converted to hydroxamic acid derivatives 12a,b by reaction with hydroxylamine. The second approach involved introducing a triazole next to the hydroxamic acid, which binds to Zn²⁺ binding at the enzymes’ catalytic centre. The synthesis was again based on Cu(I)-catalyzed azide-alkyne cycloaddition. Thus, the interaction between ferrocene-containing azides 13a,b and 15a,b and methyl propiolate led to heterocyclic esters, which were subsequently converted to the desired products.

Docking studies were performed on hydroxamic acids 12b and 16b. Compound 12b demonstrated a good fit, suggesting good inhibitory activity against HDAC8. In contrast, compound 16b exhibited steric hindrance and an inability to bind to the catalytic centre of HDAC8, as also evidenced by the biological evaluation using the IC₅₀ values against HDACs. Compounds 14 and 16a exhibited no inhibitory activity, demonstrating that chain branching adjacent to the hydroxamic acid is not sterically tolerated. However, compound 16b exhibited micromolar activities against a number of HDACs. In contrast, compound 12b showed excellent HDAC inhibition with IC₅₀ values in the nM range, even slightly lower than that of SAHA. The analogue compound 12a, with a shorter chain length, showed a weaker inhibitory activity, confirming the linker length–activity relationship. Additionally, the JAHAs were evaluated for cytotoxic activity against breast cancer cells (MCF7) and only 12b showed significant growth inhibition. Flow cytometry experiments revealed that the main mechanism of antiproliferative activity of the compound was due to induction of cell cycle arrest and subsequent cell death. To confirm the potency of 12b as an HDAC6 inhibitor, the activity of the enzyme was assayed in Xenopus laevis embryos and the click analogue showed activity comparable to that of SAHA [36].

Tang et al. designed and synthesized ferrocenyl HDACis for triple-negative cancer therapy [37]. To avoid the toxicity typically associated with the hydroxamic acid group, which is prone to hydrolysis and formation of the potentially mutagenic hydroxylamine, they developed ferrocene-containing non-hydroxamic acid derivatives. The authors were encouraged by the development of a selenocyanide-group-containing analogue of SAHA, named SelSA 17 (Figure 6), which showed reasonable zinc-binding activity and considerable potency against HDAC. It was found that replacing the phenyl ring of SelSA with a ferrocenyl group as a bioisostere resulted in increased activity. Two ferrocenyl selenocyanid derivatives were synthesized and evaluated for their biological activities: the full analogue of SelSA 18 (Fc-SelSA) and a compound in which the amide functionality was replaced by a ketone 19 (Figure 6). The inhibitory activity of the compounds against HDAC in the HeLa cell line was demonstrated using a fluorescence assay.

The synthesis was accomplished by the amidation of aminoferrocene 7 or acylation of ferrocene with 6-bromohexanoyl chloride, affording intermediates 20 and 21, respectively. These bromides were subsequently treated with potassium selenocyanide in CH₃CN to give the target compounds 18 (Fc-SelSA) and 19 (Figure 7).

A docking study of Fc-SelSA with HDAC was performed to investigate the binding mode and the results showed that the optimal binding pose in the active site pocket of the enzyme is similar to those of SelSA and SAHA. The selenocyanide group of SelSA and Fc-SelSA was shown to coordinate perfectly with the catalytic zinc ion, forming two additional hydrogen bonds that stabilize the chelation in the catalytic domain. The absence of one of these hydrogen bonds in SAHA may explain why the selenocyanide compounds had a stronger HDAC inhibitory activity. Additionally, the amide group of Fc-SelSA was found to provide two hydrogen bonds at the entrance of the catalytic pocket, whereas SAHA and SelSA did not have similar interactions, thus explaining its enhanced activity.

The results of the docking assay were supported by in vitro tests of HDAC inhibitory activity. The results showed that Fc-SelSA (IC₅₀ = 14.8 nM) outperformed the activity of SelSA (IC₅₀ = 20.3 nM). Apparently, using the ferrocenyl moiety as a cup group instead of the phenyl ring of SelSA was a successful strategy to enhance HDAC inhibition. Compound 19, in which the amide function of Fc-SelSA was replaced by a carbonyl, displayed a 26-fold lower activity (IC₅₀ = 386.0 nM), highlighting the relationship between HDAC inhibition and hydrogen bond interactions of the amide group.

The ferrocenyl derivatives were screened for their antiproliferative activity against two breast cancer cell lines MCF-7 (hormone-dependent breast cancer) and MDA-MB-231 (triple-negative breast cancer). Compared to SelSA, Fc-SelSA showed a stronger antiproliferative effect on both cell lines, with a 7-fold higher selectivity for MDA-MB-231 cells, which is strongly associated with HDAC inhibition. Again, the change in the amide function of Fc-SelSA with a carbonyl group in compound 19 resulted in a decrease in activity. In addition, the compounds were non-toxic to normal cells MCF-10A (human normal breast epithelial cells) and Vero (healthy kidney epithelial cells). An in vivo study in an MDA-MB-231-tumour-bearing mouse model was conducted with Fc-SelSA. The ferrocenyl compound showed no adverse side-effects and significantly reduced the tumour volumes, outperforming SelSA, which was used as a positive control. Thus, Fc-SelSA emerged as a promising new HDAC inhibitor for the treatment of triple-negative breast cancer [37].

Marinero et al. designed and synthesized ferrocifen-SAHA hybrids as chemotherapeutic agents against triple-negative breast cancer [38]. The strategy used was to combine several active structures into one molecule, thereby modulating different cellular pathways and demonstrating greater efficacy than a single-target drug. The authors synthesized six hybrid compounds combining tamoxifen (TAM) or ferrocifen (FcTAM) structural motifs with SAHA, and SAHA-type molecules, 8-oxo-8-(phenylamino)octanoic acid (OPOA), and N¹-phenylsuberamide (PSA), bearing carboxylic acid and primary amide functions, respectively (Figure 8).

The synthesis of the hybrids was accomplished starting from the already known anilines 28a,b, which were previously prepared in Jouen’s group by McMurry cross-coupling reactions between 4-aminobenzophenone and the corresponding ketone [39]. The subsequent reaction of the anilines with one of the activated forms of suberic acid, such as suberic anhydride, suberoyl chloride, or suberoyl ethyl carbonate, resulted in the formation of the desired FcTAM-OPOA 23 and TAM-OPOA 26, but in all cases, bisanilides were formed as unwanted by-products. Suberoyl chloride was chosen as the most suitable reagent due to its ease of formation and rapid reaction with anilines 28a,b (Figure 9). The obtained carboxylic acids 23,26 were activated as ethyl carbonates by reaction with ClCO₂Et and NEt₃ and subsequently reacted with freshly prepared hydroxylamine to produce FcTAM-SAHA 22 and TAM-SAHA 25, or with sodium amide to give FcTAM-PSA 24 and TAM-PSA 27 [38].

The compounds were evaluated for their antiproliferative activity against triple-negative MDA-MB-231 and hormone-dependent MCF-7 breast cancer cell lines. The hybrid FcTAM-SAHA demonstrated better cytotoxicity (IC₅₀ = 0.7 μM) against MDA-MB-231 compared to its two parent molecules ferrocifen (IC₅₀ = 2.6 μM) and SAHA (IC₅₀ = 3.6 μM) alone. The organic analogue TAM-SAHA was less cytotoxic (IC₅₀ = 8.6 μM). The results in MCF-7 cells showed that FcTAM-SAHA (IC₅₀ = 2.0 μM) is more active than ferrocifen (IC₅₀ = 4.4 μM) and TAM-SAHA, but less potent than SAHA (IC₅₀ = 1.0 μM). The amide analogue FcTAM-PSA also showed better antiproliferative activity than both ferrocifen and SAHA in MDA-MB-231 cells (IC₅₀ = 0.5 μM) but again worse than that of SAHA in MCF-7 cells (IC₅₀ = 1.8 μM)). In addition, the compounds were examined for their ability to inhibit HDACs, and the importance of the hydroxamic function for HDAC inhibition was undoubtedly proven. Interestingly, FcTAM-PSA did not possess HDAC activity, but nevertheless demonstrated strong antiproliferative effects on the cell lines, which undoubtedly points to a novel mechanism of action of the organometallic compounds. The authors speculated that the redox properties of ferrocene compounds and the production of reactive oxygen species may explain their astonishing activity against the two breast cancer cell lines [38,40].

In a subsequent paper, Marinero et al. extended their research towards ferrocenyl compounds bearing chains of different lengths [41]. They synthesized analogues of FcTAM-SAHA and FcTAM-PSA with chain lengths of two and four methylene groups by the initial reaction of 28a with succinyl chloride (ClCO(CH₂)₂COCl) and adipoyl chloride (ClCO(CH₂)₄COCl), respectively, and the compounds were evaluated against MDA-MB-231 and MCF-7 breast cancer cells. The results showed that the chain shortening did not dramatically affect the biological activity, although the two cell lines had different sensitivities to the compounds. The primary amides showed similar antiproliferative effects against both types of cancer cell lines and, interestingly, the succinamide was the most active compound against MCF-7 cells. However, the cytotoxicity of the hydroxamic acid derivatives decreased with the chain length shortening, suggesting that their activity is related to HDAC inhibition.

In 2014, Marinero et al. synthesized hydroxytamoxifen-SAHA 29 and hydroxyferrocifen-SAHA 30 hybrids (Figure 10) and studied their cytotoxicity against triple-negative MDA-MB-231 and hormone-dependent MCF-7 breast cancer cells [42]. The work can be seen as a continuation of the development of organometallic analogues of hydroxytamoxifen started by the Jaouen group in 1996 [43]. The hydroxyferrocifen (FcOHTAM) demonstrated significant cytotoxicity on both hormone-dependent and hormone-independent breast cancer cell lines and showed remarkable antiproliferative effects against MDA-MB-231, while hydroxytamoxifen (OHTAM) was inactive under the same conditions [44]. Since combining the structures of ferrocifen and SAHA into a single hybrid FcTAM-SAHA led to an increase in antiproliferative effects, the authors became interested in developing FcOHTAM-SAHA hybrids. The synthesis of the hydroxyferrocifen hybrids was carried out by procedures similar to those used for the ferrocifens [38,42]. The cytotoxic activities of the organometallic compounds were better than those of their organic analogues. FcOHTAM-SAHA showed the best performance against MDA-MB-231 cells (IC₅₀ = 1.3 μM), a value lower than that of its organic analogue (IC₅₀ = 5.2 μM) and SAHA (IC₅₀ = 3.6 μM). The antiproliferative effect of FcOHTAM-SAHA (IC₅₀ = 1.5 μM) against MCF-7 cells was much higher than that of OHTAM-SAHA (IC₅₀ = 10.9 μM), but in the range of SAHA (IC₅₀ = 1.0 μM) and the non-phenolic hybrid FcTAM-SAHA (IC₅₀ = 2.0 μM). This suggested that the hydroxyl group did not improve the cytotoxicity and raised the question of the mechanism of action of the organometallic compounds.

In 2016, Li et al. reported ferrocenyl compounds based on a bicyclic core skeleton as dual-acting estrogen receptor (ER) and HDAC inhibitors for breast cancer therapy [45]. The authors implemented the strategy of developing bivalent agents by combining two bioactive structures in one molecule. They stepped on their previous discovery of ER ligands based on 7-oxabicyclo[2.2.1]hept-5-ene core and, in particular, the successful combination of 5.6-bis-(4-hydroxyphenyl)-7-oxabicyclo[2.2.1]hept-5-ene-2-sulfonic acid (OBHS) in combination with HDACi in a dual-acting OBHS-HDACi conjugate [46]. The combination of the latter with the ferrocenyl moiety in FcOBHS-HDACis was considered a possibility to expand the activity of the conjugates towards both ER(+) and ER(−) breast cancer cells. The ferrocene was introduced into the OBHS-HDACi backbone by attachment to the phenyl ring of the sulfonate unit, or by replacement of a phenol ring at C6 with a ferrocenyl group.

The synthesis of conjugates 33a–i and 36 was performed by the Diels–Alder cycloaddition of 3,4-disubstituted furan derivatives with different dienophiles (Figure 11). The starting furan derivatives 31a–c and 34 and the vinyl sulfonate 35 were synthesized according to previously developed protocols [46]. The synthesis of the ferrocenyl vinyl sulfonates 32a–c was achieved by Friedel–Crafts acylation or the alkylation of ferrocene with 3-methoxybenzoyl chloride, 4-methoxybenzoyl chloride, or 4-methoxybenzyl alcohol, followed by O-demethylation with BBr₃ and a final reaction of the phenols with 2-chloroethanesulfonoyl chloride in the presence of NEt₃ [45]. The most important Diels–Alder reactions were carried out with good yields and stereoselectivity towards the exo products [47].

The FcOBHS-HDAC conjugates demonstrated good ER binding affinities with selectivity for ERα, where compounds 33b,e,h with a chain length of six methylene groups demonstrated a better affinity. These compounds also showed excellent Erα antagonist activity. Some of the FcOBHS-HDAC conjugates exhibited inhibitory activity against HDAC1 and HDAC6 with the suberic acid conjugate 33e (n = 6, X = 4-CO) showing the best inhibitory activity against HDAC1 with an IC₅₀ value of 0.197 µM, which unfortunately is much lower than the activity of the reference compound SAHA (IC₅₀ = 0.0474 µM). The conjugates were also screened against MCF-7, MDA-MB-231, and DU-145 cancer cell lines, demonstrating to be more potent against the hormone-independent ER(−) breast cancer cells (MDA-MB-231). Furthermore, all of the conjugates were found to be non-toxic to the healthy Vero cell line. The most promising compound was again 33e, which demonstrated the highest activity against both hormone-dependent and -independent cancer cells. Thus, it was again demonstrated that the incorporation of a ferrocene moiety in the structure of a known anticancer agent is enhancing the activity against ER(−) breast cancer cells.

In 2017, Ocasio et al. reported a ferrocene-containing o-aminoanilide analogue of JAHA, named Pojamide, which was characterized with HDAC3 inhibition and remarkable redox-triggered ferrocenium activity in cells [48]. Aware of the documented toxicity of the hydroxamate zinc-binding group and the need for selective HDACis, the authors modified the structure of TCH106, an o-aminoanilide analogue of SAHA and selective HDAC3 inhibitor, with a ferrocenyl moiety. By the reaction of 6-oxo-6-(ferrocenylamino)hexanoic acid 37a or 8-oxo-8-(ferrocenylamino)octanoic acid 37b with N-Boc-o-phenylenediamine and propanephosphonic acid anhydride (T3P) as a coupling reagent and subsequent removal of the Boc-protection (Figure 12), two compounds with different linker-lengths were synthesized, namely 38a and 38b (Pojamide).

A docking study of 38a and 38b with HDAC3 was performed to investigate the binding mode, and the results showed that both compounds were bound to the zinc active site but in a different manner. Compared to Pojamide, the optimal binding pose of the shorter chain 38a lacks two of the hydrogen bonds and coordinates to zinc with its benzamide carbonyl instead of the aniline–zinc interaction typical for o-aminoanilides. In addition, the docking analysis suggested that the ferrocenyl substituent in Pojamide was responsible for HDAC3 selectivity due to the collision of the bulky organometallic moiety with two amino acids at the entrance to the active site of HDAC8. The selective binding of Pojamide to HDAC3 was confirmed using Xenopus laevis embryos as a model system. The levels of histone H4K12, which is a target of HDAC3, increased in embryos treated with Pojamide. The compound was also tested against a panel of HDACis, demonstrating selectivity toward HDAC3 over HDAC1,2,4–8. The activity of Pojamide was evaluated in cell lines overexpressing HDAC3, namely HeLa, HT-29, and HCT116 [49,50]. The antiproliferative effect of Pojamide against HCT116 was approximately four times lower than that of the control TCH106, suggesting that combined HDAC1,3 inhibition is desirable for optimal anticancer activity. Finally, the authors explored the Pojamide-induced redox pharmacology by using it in combination with sodium nitroprusside (SNP) and glutathione (GSH), which induced the generation of Fe(III) species and led to greatly enhanced cytotoxicity and DNA damage.

In 2019, Luparello et al. reported additional studies on the activity of Pojamide, its shorter analogue 38a, and a ferrocenium tetrafluoroborate analogue against triple-negative MDA-MB-231 breast cancer cells [51]. Despite the differences in their HDAC inhibitory activity demonstrated in the previous study, both Fe(II) compounds 38a,b demonstrated equipotent antiproliferative activity. The Fe(III) tetrafluoroborate derivative exerted no effect on cell proliferation. The cytotoxic effect of the two ferrocenyl compounds was linked to the induction of oxidative stress, via increased ROS generation, impairment of mitochondrial function, and modulation of the autophagy process, manifested by a reduction in the number and size of the acidic vesicular organelles (AVOs) [51,52].

In 2023, Yan et al. published the synthesis and biological evaluation of a library of ferrocene- and ferrocenophane-based hydroxamic acids as selective HDAC6 inhibitors [53]. The N-hydroxybenzamide, previously reported as being part of the structures of HDAC6-selective inhibitors, was chosen as the zinc binding group. This is explained by the larger pocket on the active side of HDAC6, which allows coordination with sterically bulky groups [54]. The authors used ferrocenes, [3]-ferrocenophanes, and 2-aza-[3]-ferrocenophanes as cap groups, and connected them to the benzohydroxamic acid via a variety of linkers to create a wide range of ferrocenyl hydroxamic acids. The most active compounds were the ferrocenyl derivative with a piperazine linker 39, the ferrocenecarboxamide with a single methylene group linker 40, and their [3]-ferrocenophane and 2-aza-[3]-ferrocenophane structural analogues, 41 and 42, respectively (Figure 13).

The synthesis of the ferrocenyl derivative 39 was accomplished by an initial reductive amination of ferrocenecarboxaldehyde 43 with piperazine to give amine 44, which was then reacted with methyl 4-(bromomethyl)benzoate. The ester intermediate was converted to hydroxamic acid when treated with NH₂OH in MeOH (Figure 14). Ferrocenecarboxamide 40 was synthesized by the amide coupling of ferrocenecarboxylic acid 45 with methyl 4-(aminomethyl)benzoate and the final transformation of the ester into hydroxamic acid. The rigid [3]-ferrocenophane structure of 41 was built by the implementation of tandem Friedel–Crafts acylation and the alkylation of ferrocene with acryloyl chloride at −78 °C in the presence of AlCl₃. The resulting cyclic ketone 46 was reduced to the alcohol 47, which was acylated and subjected to nucleophilic substitution with piperazine to form the amine 48. The latter was reacted with methyl 4-(bromomethyl)-3-fluorobenzoate and the resulting ester intermediate was converted into 41. The 2-aza-[3]-ferrocenophane derivative 42 was obtained by the cyclic reductive amination of 1,1′-ferrocenedicarboxaldehyde 49 with methyl 4-(aminomethyl)benzoate in the presence of NaBH(OAc)₃ and the subsequent transformation of the intermediate ester 50 into the corresponding hydroxamic acid (Figure 14).

A library of 18 ferrocene- and 9 ferrocenophane-based hydroxamic acids was evaluated for HDAC6 inhibitory activity with JAHA and SAHA as reference compounds. From the ferrocene series, compounds 39 and 40 displayed the best, but still moderate, activity, both with IC₅₀ values of 212 nM, approximately 10 times lower than that of SAHA. The replacement of the ferrocene with [3]-ferrocenophane resulted in increased inhibitory activities with the best result demonstrated by compound 41 (IC₅₀ = 38.2 nM), which was developed after fine-tuning the initial structure. The 2-aza-[3]-ferrocenophane 42 also exhibited relatively good HDAC6 inhibitory activity with an IC₅₀ value of 76.5 nM. In addition, selected compounds were tested for HDAC6 selectivity over other HDAC subtypes. Compound 41 demonstrated at least a 67-fold selectivity for HDAC6 over HDAC1–5,7,8 and 42 showed an approximately 50-fold selectivity over HDAC1–3. The authors suggested that the unique structural features of the ferrocenophanes are likely to be a good match for the cap region of HDAC6.

The antiproliferative effect of compounds 41 and 42 was evaluated in a panel of cancer cell lines, including 22RV1 (prostate cancer), MM1.S (multiple myeloma), MV4–11 (monocytic leukemia), JEKO-1 (mantle cell lymphoma), and 4T1 (breast cancer). Both compounds demonstrated moderate antiproliferative activity in all cell lines, but lower than that observed for the reference JAHA, which was explained by the selectivity for HDAC6 over the other enzyme isoforms. The isoform selectivity of compound 41 was also confirmed by the dose-dependent accumulation of acetylated α-tubulin while having a negligible effect on the level of acetylated histone H3 [55]. Two probable mechanisms referring to apoptosis were proposed for compounds combining in their structures ferrocene and hydroxamic acid to induce programmed cell death. One involves triggering the intrinsic pathway, typical for HDAC inhibitors. The other involves ferrocene-induced ROS overproduction [53,56]. An increase in the ROS levels was detected after the incubation of 4T1 cells with 41 for 1 h [53]. These results further support the importance of the ferrocenyl moiety in the development of new antiproliferative agents.

3. Conclusions

In this mini-review, we summarized the development of HDAC inhibitors with a ferrocenyl moiety as anticancer agents. The review started with the pioneering work of Spencer and the creation of JAHA, and the variety of its structural modifications including the photoreleasable p-Fc-SAHA, the triazole-containing click JAHA-analogues, and the o-aminoanilide Pojamide. It also covered the effectiveness for triple-negative cancer therapy of ferrocenyl-selenocyanides, the Ferrocifen-SAHA multifunctional agents developed by Jaouen’s group, the sulfonates with a bicyclic core, and the recently published ferrocenophanes. The exploration of this specific and, in some respects, limited topic illustrated how the combination of one lead compound, SAHA, with ferrocene can enhance the existing inhibitory activity and give a rise to a variety of additional biological activities. The introduction of the bulky, lipophilic, barrel-shaped organometallic moiety obviously influenced the specific binding to the biological targets and additionally took advantage of its reversible redox properties and ROS production. Special attention was given to the compound’s synthesis, highlighting the diversity of possible ferrocene functionalizations, leading to structural modifications and opening up the opportunity of SAR discussions. Combining ferrocene with the molecules of known HDACis led to the discovery of hybrids with improved anticancer activity, particularly against triple-negative breast cancer models. The main objective of this mini-review was again to draw the attention of the scientific community to the significant potential of this organometallic compound in medicinal chemistry.