1. Introduction
Histone deacetylases (HDACs) are a group of epigenetic enzymes that regulate gene expression without altering the underlying DNA sequence. They function by removing acetyl groups from lysine residues on both histone and non-histone proteins, thereby modulating intracellular acetylation levels in concert with histone acetyltransferases [ 1 , 2 ]. In mammals, 18 HDAC isozymes have been identified and are classified into four groups based on their sequence similarity to yeast orthologues: Class I (HDAC1, HDAC2, HDAC3, HDAC8), Class IIa (HDAC4, HDAC5, HDAC7, HDAC9), Class IIb (HDAC6, HDAC10), Class III (sirtuins 1–7), and Class IV (HDAC11).
HDAC6 is the largest member of the HDAC family, comprising 1215 amino acids. It is characterized by two tandem catalytic domains (CD1 and CD2), located in the N-terminal and central regions, respectively [ 3 ]. Its C-terminal region contains a zinc finger ubiquitin-binding domain (ZnF-UBP), which mediates interactions with the ubiquitin-proteasome and aggresome pathways, thereby contributing to the clearance of misfolded proteins [ 4 ].
In 2016, Hai and Christianson resolved the crystal structure of HDAC6’s catalytic domains CD1 and CD2 using X-ray crystallography [ 5 ], providing crucial insights into its structure and function. Their findings showed that CD2 possesses broader substrate specificity and greater catalytic activity than CD1, primarily due to distinct structural features within their active sites. This broader specificity refers to enzymological studies showing that CD2 can act on a variety of peptide substrates, whereas CD1 exhibits much narrower specificity. The catalytic Zn 2+ ion in CD2 plays a pivotal role in ligand recognition, with monodentate coordination being particularly favorable for specific inhibitor classes [ 6 , 7 , 8 ].
HDAC6 also contains a serine/glutamate-rich tetradecapeptide repeat domain that facilitates interactions with microtubule-associated proteins such as Tau, thereby contributing to the regulation of protein aggregation [ 9 ]. Misfolded proteins have a strong tendency to aggregate, thereby disrupting essential cellular processes, and the identification of the aggresome has significantly advanced our understanding of the cellular mechanisms involved in their recognition, processing, and degradation. Aggresomes form in response to the ectopic expression of various misfolded proteins and serve as central depots for aggregated, polyubiquitinated protein species, and HDAC6 plays a pivotal role in this pathway, being selectively involved in aggresome formation through its interaction with polyubiquitinated proteins. At the molecular level, HDAC6 simultaneously binds to ubiquitinated misfolded proteins and the motor protein dynein. This dual interaction facilitates the retrograde transport of protein aggregates along microtubules toward the microtubule-organizing center, where aggresomes form. Disruption of HDAC6 function impairs the delivery of these polyubiquitinated proteins to dynein, compromises aggresome assembly, and hampers the clearance of misfolded proteins from the cytoplasm. Consequently, cells deficient in HDAC6 exhibit heightened sensitivity to proteotoxic stress, ultimately triggering apoptotic cell death [ 10 ]. Due to its unique structural characteristics and broad substrate specificity, HDAC6 plays other multifaceted roles in various cellular processes. It regulates cytoskeletal dynamics via tubulin deacetylation, modulates cellular stress responses through the promotion of autophagy, and influences immune signaling by deacetylating key mediators, such as those involved in the NF-κB pathway, thereby affecting inflammation and immune regulation. Additionally, HDAC6 indirectly modulates gene expression by modifying non-histone proteins involved in transcriptional control [ 9 ].
These diverse functions make HDAC6 essential for maintaining cellular homeostasis and responding to stress while also establishing it as a compelling therapeutic target in diseases such as cancer, neurodegeneration, and inflammation. Recent studies have shown that HDAC6 overexpression is associated with increased cell proliferation, enhanced migratory capacity, poor clinical prognosis, and the development of drug resistance across various cancer types [ 11 , 12 ]. Given its distinctive structural and functional roles, HDAC6 has emerged as a promising target for therapeutic intervention. Selective inhibition of HDAC6 has demonstrated notable potential in attenuating tumor progression, reducing metastasis, and overcoming drug resistance in a range of preclinical cancer models [ 13 , 14 ].
Currently approved HDAC inhibitors (HDACis), such as vorinostat and panobinostat, exhibit limited isozyme selectivity, often inhibiting both Class I and Class II HDACs. This broad-spectrum activity affects multiple cellular pathways and is associated with significant adverse effects [ 15 , 16 ]. Consequently, there is growing interest in developing selective HDAC6 inhibitors. Several candidates are currently under preclinical and clinical investigation. Notably, some hydroxamic acid-based compounds have demonstrated high selectivity for HDAC6 and the ability to cross the blood–brain barrier, exhibiting neuroprotective effects in various disease models [ 17 ].
Furthermore, Lv et al. reported a new class of brain-penetrant, non-hydroxamate HDAC6-selective inhibitors designed to minimize the genotoxicity commonly associated with hydroxamate-based scaffolds [ 18 ].
Although several pan-HDAC inhibitors have been approved by the FDA, they are currently restricted mainly to the treatment of cutaneous T-cell lymphoma [ 19 ]. More recent preclinical research has focused on the potential of HDAC6 inhibitors, particularly in combination therapies for cancers such as melanoma, glioblastoma, and various hematological malignancies [ 20 , 21 , 22 , 23 ], including multiple myeloma (MM), a still incurable plasma cell disorder characterized by abnormal proliferation of malignant plasma cells in the bone marrow [ 24 , 25 ].
Multiple myeloma (MM) represents the second most common hematologic malignancy worldwide and is defined by the clonal expansion of neoplastic plasma cells within the bone marrow microenvironment. Its pathogenesis involves a multifactorial process driven by intricate genetic and epigenetic alterations [ 24 , 26 , 27 ], including chromosomal translocations, copy number variations, and dysregulation of critical signaling pathways [ 28 , 29 , 30 ], which collectively promote cellular proliferation, survival, and resistance to apoptosis.
Despite substantial therapeutic advancements—such as the incorporation of proteasome inhibitors, immunomodulatory agents, monoclonal antibodies [ 31 ], immune-based therapies, and a growing array of emerging targeted agents—MM remains an incurable malignancy for the majority of patients. Most individuals ultimately experience disease relapse and progression, frequently accompanied by increasing resistance to available therapeutic modalities [ 32 , 33 , 34 , 35 , 36 ]. Moreover, complications such as peripheral neuropathy can significantly impact patients’ quality of life and limit therapeutic options [ 37 ].
In this regard, HDAC6 inhibitors have demonstrated encouraging anti-MM activity, mainly due to their capability to disrupt the aggresome-autophagy pathway and induce apoptosis in MM cells. This pro-apoptotic effect increases when using these inhibitors in combination with proteasome inhibitors. In addition to their cytotoxic properties, HDAC6 inhibitors enhance antitumor immune responses, contributing further to their therapeutic potential in MM [ 38 ]. According to ClinicalTrials.gov, two HDAC6 inhibitors ( ACY-1215 and ACY-241) have advanced into clinical trials for MM therapy. Despite their low nanomolar half-maximal inhibitory concentration (IC 50 ) values against HDAC6, both compounds exhibit only modest selectivity, with approximately 10-fold higher affinity for HDAC6 over class I isozymes HDAC1, HDAC2, and HDAC3 [ 39 ]. Given these limitations, and the well-established role of HDAC6 in myelomagenesis, recent research efforts have been directed toward the development of next-generation HDAC6 inhibitors with enhanced potency and selectivity. Notably, a novel irreversible inhibitor incorporating a phenylsulfonylfuroxan-based hydroxamate scaffold has shown improved anti-MM activity [ 38 ], underscoring ongoing advancements in this therapeutic area.
Based on these insights, we performed a Ligand-Based Pharmacophore Screening (LBPS) of commercially available compound libraries, using a structurally diverse set of known HDAC6 inhibitors with established anticancer activity. Subsequent molecular docking and molecular dynamic simulations led to the identification of three promising candidates as potential HDAC6 inhibitors. Among them, AKOS030273637 ( Compound 10 ), featuring a benzyl 4-[4-(hydroxyamino)-4-oxobutylidene]piperidine-1-carboxylate scaffold, was confirmed to inhibit HDAC6 enzymatic activity, with a significant anti-tumor effect against MM cell lines, including those resistant to standard therapies.
2. Results
2.1. Identification of Potential HDAC6 Inhibitors via Virtual Screening
In this study, the computational pipeline was launched by downloading and preparing the HDAC6 isozyme structure (PDB ID: 5EDU) [ 40 ], along with compound libraries from both MolPort Natural Compounds [ 41 ] and Asinex (BioDesign Library and Elite Library) [ 42 ], which comprise a total of over 300,000 molecules. To minimize stereochemical ambiguity and circumvent issues related to enantiomeric purity, the database was filtered based on molecular chirality, retaining only achiral compounds. This step refined the screening library to 142,486 entries. Subsequently, a ligand-based pharmacophore model was generated using a training set of potent HDAC6 inhibitors with different structural motifs ( Table 1 and Table S1 ).
Among the ten generated pharmacophore hypotheses ( Table S2 ), the top-ranked model—featuring two hydrogen bond acceptors (HBA), one hydrogen bond donor (HBD), one hydrophobic feature (NI), and one negatively ionizable group (HY)—exhibited strong predictive performance, achieving an area under the curve (AUC) of 87.0% and a sensitivity of 75% in virtual screening assays ( Figure 1 A). This model served as an initial screening tool to prioritize compounds for subsequent structure-based analysis.
More in detail, the final pharmacophore model is defined by two HBAs and one HBD, all spatially localized within the hydroxamic acid moiety, a functional group conserved across all training compounds and known to play a pivotal role in zinc coordination within HDAC active sites. A HY feature is also present and corresponds to different aromatic systems that contribute to ligand-enzyme interactions via π–π stacking and hydrophobic contacts. These include: the phenyl ring of the dihydroquinoline scaffold in Compound 2 , the phenyl moiety of the N-phenylaniline group in Compound 6 , the naphthalene ring system in Compound 1 , the phenyl ring fused to the indole core in Compound 3 , the pyridine ring in Compound 4 , and the pendant phenyl substituent on the indole system in Compound 5 ( Figure S1 ).
Notably, the model does not incorporate a defined linker region, which is typically observed in classical HDAC inhibitor pharmacophores bridging the cap group and the zinc-binding domain. This absence could be attributed to the structural heterogeneity of the training set, prompting the generation of a consensus pharmacophore based solely on features that are universally conserved across all compounds. Pharmacophore validation was conducted through the virtual screening of the curated test set libraries, followed by receiver operating characteristic (ROC) curve analysis. The model demonstrated high discriminative power, correctly identifying 9 out of 12 active compounds, with a sensitivity of approximately 75% and an area under the ROC curve (AUC) of 0.87 ( Figure 1 B). This validated pharmacophore model was subsequently utilized to filter the prepared compound libraries, prioritizing candidates based on optimal pharmacophore fit scores. This refinement step yielded a subset of 256 compounds—comprising 101 from MolPort, 26 from Asinex BioDesign Library, and 129 from Asinex Elite Library. These selected hits were then subjected to a structure-based virtual screening (SBVS) workflow to further assess their binding potential within the HDAC6 catalytic site ( Figure 2 ).
To establish a reliable threshold for the subsequent SBVS, we performed re-docking of the co-crystallized ligand Trichostatin A (TSA) into both chain A and chain B of the prepared HDAC6 model. Chain A was selected for further docking studies as its re-docked pose yielded the lowest root-mean-square deviation (RMSD) of 1.12 Å, confirming the accuracy and reproducibility of the docking protocol. The Glide Score (G-Score) of Trichostatin A ’s best pose in chain A was used as a cut-off criterion. A total of 19 compounds demonstrated better G-Scores than Trichostatin A .
These 19 hits were subjected to fingerprint-based clustering using the Tanimoto coefficient. The optimal number of clusters was determined with Canvas software (version 3.9) [ 44 ] employing the Kelley criterion, resulting in five distinct clusters. For each cluster, the representative compound was identified by selecting the molecule with the highest number of shared features relative to other cluster members, based on centroid comparison ( Figure S2 ). Among the five identified molecules, only four were commercially available from the same supplier. As a result, AKOS030314251 was excluded, and the remaining four were selected for further analysis ( Figure 2 , Table 2 ).
The four selected hit compounds exhibit distinct chemical scaffolds and incorporate structural features consistent with HDAC6 inhibition. Compound 7 (AKOS030496586) integrates phenolic, carboxylic, and enone functionalities, allowing for a combination of polar interactions and extended π-conjugation within the active site. The second hit , Compound 8 or AKOS000531201, possesses a fused heterocyclic system capable of forming multiple hydrogen bonds and π–π stacking interactions, along with a methoxy-substituted aromatic ring likely to engage hydrophobic regions of the binding pocket. Compound 9 (AKOS030461429) features a thiazole-triazole core and a hydrophobic isopropoxyphenyl moiety, suggesting potential for both metal coordination and hydrophobic interactions within the HDAC6 active site. Finally, Compound 10 (AKOS030273637) contains a hydroxamic acid group—a well-established zinc-binding motif in HDAC inhibitors—linked to a piperidine scaffold, which may contribute to improved binding specificity and physicochemical properties.
Thus, similar to Trichostatin A , all selected compounds possess functional groups capable of coordinating the catalytic zinc ion within the HDAC6 active site. Also, Compound 7 establishes a π–π interaction with Phe620 and a hydrogen bond with His651, supporting a combination of polar and aromatic interactions. Compound 8 primarily interacts through π–π stacking with His500 and Phe620, indicating a predominantly hydrophobic binding profile. Compound 9 is predicted to engage in three π–π stacking interactions with His611, His651, and Phe680, stabilizing its binding conformation. Finally, Compound 10 forms hydrogen bonds with His610 and Tyr782, along with a π–π interaction involving His500, suggesting a strong and multifaceted binding mode. These interaction patterns are illustrated in Figure 3 .
To gain deeper insights into the molecular recognition mechanisms of HDAC6, molecular dynamic simulations (MDs) were conducted in explicit water on complexes of the four selected hit compounds. As a control, molecular dynamics (MD) simulations were also performed on Trichostatin A to validate the simulation protocol. TSA is a well-characterized broad-spectrum (pan-HDAC) inhibitor, exhibiting markedly higher potency against class I HDAC isoforms, such as HDAC1, than against class IIb enzymes like HDAC6.
The RMSD plots of the heavy atoms of the protein for all protein–ligand complexes, including Trichostatin A and the selected hit compounds, are shown in Figure S3 . This figure also includes a table summarizing the average RMSD values for each system, providing a clear overview of the protein’s structural stability throughout the simulations.
Trichostatin A demonstrated stable binding throughout the simulation, as indicated by consistently low RMSD values of ligand heavy atoms, calculated by aligning trajectory frames to the protein backbone of the initial structure ( Figure 4 A). Additionally, it maintained coordination with the catalytic zinc ion over the full simulation time ( Figure S4 ), confirming the robustness of the simulation setup.
Among the four hit compounds, Compound 8 , Compound 9 and Compound 10 exhibited stable binding profiles, maintaining low RMSD values and sustained interactions with the catalytic zinc ion ( Figure 4 A,B). In contrast, Compound 7 initially remained bound, but during the final 50 ns of the simulation, it dissociated from the active site and lost zinc coordination ( Figure 4 A,B).
Detailed interaction analyses revealed that Compound 8 formed a stable π–π interaction with Phe620, Compound 10 was involved in a hydrogen bond with His610 and a water bridge with Asp742, and Compound 9 established two water bridges with His610 and His652 ( Figure 4 B).
To complement the structural stability analysis, Molecular Mechanics Generalized Born Surface Area (MM-GBSA) calculations were performed in order to estimate binding free energies (ΔG bind ) for the three stably bound compounds and Trichostatin A . While MM-GBSA offers a valuable approximation of binding affinity, it predominantly reflects enthalpic contributions and does not explicitly account for entropic factors, which can also significantly influence ligand binding. Trichostatin A displayed a mean ΔG bind of −28.35 kcal/mol. Notably, Compound 10 exhibited a more favorable binding energy of −45.41 kcal/mol, while Compounds 9 and 8 showed values of −18.82 kcal/mol and −22.25 kcal/mol, respectively ( Figure 4 C).
Based on these results, all three compounds were selected for purchase and subjected to subsequent experimental validation.
2.2. Fluorometric Validation of HDAC6 Inhibition: Compound 10 Emerges as a Potent Lead Compound
The ability of Compound 8 , Compound 9 , and Compound 10 to inhibit HDAC6 enzymatic activity was assessed using a fluorometric assay designed to measure HDAC6-specific deacetylation activity. The results, summarized in Table 3 , demonstrate that Compound 10 was the most potent among the three compounds, exhibiting an IC 50 in the low micromolar range.
Notably, the IC 50 value of Compound 10 was significantly lower compared to those of Compound 8 and Compound 9 , indicating a stronger inhibitory effect on HDAC6 ( Table 3 ; Figure S5 ). These findings are consistent with prior in silico predictions and support the selection of Compound 10 as the lead compound for further biological validation and mechanistic studies.
2.3. Compound 10 Inhibits Multiple Myeloma (MM) Cell Viability Enhancing Acetylated α-Tubulin Expression
Next, we investigated the in vitro anti-tumor activity of Compound 10 in MM cell lines, including those that are either sensitive or resistant to standard-of-care treatments, such as proteasome inhibitors (PIs). Notably, treatment with Compound 10 resulted in a significant reduction in cell viability across all tested MM cell lines, exhibiting half-maximal inhibitory concentration (IC 50 ) values in the micromolar range, thereby demonstrating potent cytotoxic effects. Importantly, the IC 50 values were approximately two-fold lower in PI-sensitive cell lines compared to PI-resistant counterparts ( Figure 5 ), suggesting that Compound 10 is more effective in treatment-naïve or less refractory MM contexts.
Compared to the reference HDAC6-selective compound ACY1215, Compound 10 exhibited higher IC 50 ; moreover, similar to Compound 10 , ACY-1215 also displayed greater activity in PI-sensitive cell lines compared to their resistant counterparts ( Figure S6 ). Interestingly, Compound 10 elicited negligible effects on the viability of non-transformed cells (293T) or healthy peripheral blood mononuclear cells (PBMCs), suggesting a favorable therapeutic window ( Figure S7 ).
We next evaluated whether HDAC6 inhibition mediated by Compound 10 could translate into increased acetylation of known HDAC6 substrates, such as α-tubulin. Similarly to the HDAC6-selective compound ACY-1215 ( Figure S8 ), treatment with Compound 10 led to a marked increase in acetylated α-tubulin levels in MM cell lines, including both drug-sensitive and drug-resistant models ( Figure 6 ).
This increase in tubulin acetylation provides a functional readout of HDAC6 inhibition and confirms that Compound 10 effectively engages its target in a cellular context. Notably, the effect was observed regardless of the cells’ sensitivity to proteasome inhibitors, indicating that Compound 10 can likely modulate HDAC6 activity across diverse MM phenotypes.
In comparison to its effect on acetylated α-tubulin, the impact on histone H4 acetylation was moderate and even lower than that induced by ACY-1215 , further supporting the HDAC6-inhibitory activity of this compound ( Figure S9 ).
3. Discussion
HDAC6 is a unique member of the histone deacetylase family, distinguished by its structural features, including two catalytic domains and a zinc finger ubiquitin-binding domain, as well as its broad substrate specificity. HDAC6 is therefore involved in a variety of cellular processes, such as protein degradation, cell motility, stress responses, and autophagy. Consequently, it plays a pivotal role in maintaining cellular homeostasis and modulating the response to diverse forms of stress. Its involvement in multiple oncogenic and neurodegenerative pathways has positioned HDAC6 as a highly promising therapeutic target in conditions such as cancer, neurodegeneration, and inflammation.
In recent years, the selective inhibition of HDAC6 has garnered considerable interest as a therapeutic strategy to attenuate tumor progression, suppress metastasis, and overcome drug resistance—particularly in preclinical models of cancer. Given its therapeutic potential, significant efforts have been directed toward the identification of novel, selective inhibitors. Among these, pharmacophore-based virtual screening has emerged as a valuable approach for discovering new scaffolds with preferential affinity for HDAC6 [ 45 , 46 , 47 , 48 ].
Most HDAC inhibitors (HDACis), including those targeting HDAC6, incorporate a hydroxamate moiety as the zinc-binding group (ZBG), which chelates the catalytic Zn 2+ ion within the enzyme’s active site [ 49 , 50 ]. However, because this functional group is common to inhibitors of various zinc-dependent HDAC isozymes, structural selectivity must be conferred through modifications to other regions of the molecule—namely, the linker and capping groups.
In our computational drug discovery workflow, we employed a multitiered in silico approach, combining ligand-based pharmacophore modeling, molecular docking, molecular dynamic simulations, and MM-GBSA binding free energy calculations. For this initial proof-of-concept study, we deliberately focused on hydroxamate-based inhibitors, given their high prevalence in public datasets and well-characterized zinc-chelating interactions with HDAC active sites. This chemically homogeneous subset enabled the generation of a consistent and reliable pharmacophore model, leveraging the conserved binding mode of hydroxamic acids to produce a robust and predictive hypothesis.
We acknowledge that this focus represents a limitation in terms of chemical diversity. While including alternative zinc-binding groups (ZBGs) could enhance the model’s generalizability, their incorporation would require a significantly larger and more balanced training set to maintain model accuracy and avoid introducing structural noise. Nevertheless, the present model provides a strong foundation that can be systematically expanded in future studies to accommodate a broader range of ZBGs within a more comprehensive virtual screening strategy.
This approach led to the identification of three promising candidate compounds. Among them, Compound 10 exhibited the most favorable binding energy profile, outperforming the reference inhibitor Trichostatin A . This compound, abenzyl 4-[4-(hydroxyamino)-4-oxobutylidene]piperidine-1-carboxylate, was sourced from the Asinex BioDesign Library and has a molecular weight of 318.37 g/mol. According to SwissADME predictions, it displays favorable drug-likeness and medicinal chemistry properties [ 51 ]. Structurally, it conforms to the classical HDAC inhibitor architecture, comprising a ZBG, linker, and capping group. Importantly, it avoids the use of bulky capping moieties, which are often associated with reduced ligand efficiency and off-target activity. Interestingly, structurally related compounds have previously been identified as HDAC inhibitors. In detail, various alkyl piperidine hydroxamic acids exhibited submicromolar inhibitory activity. Among these, the urea analogs were the most potent, although sulfonamides and amides were also investigated [ 52 , 53 ]. Compound 10 differs from this series of alkyl piperidine hydroxamic acids by the presence of an unsaturation within the aliphatic chain, which reduces molecular flexibility. Moreover, it features a carbamate moiety instead of the urea, sulfonamide, or amide groups previously described.
Of particular note is its ability to form a stable hydrogen bond with His610, a residue located near the catalytic zinc ion. This interaction resembles the dual coordination mechanism observed in TO-317, a highly selective HDAC6 inhibitor that engages both Zn 2+ and His614 in Danio rerio HDAC6 [ 54 ]. While these findings suggest a potentially favorable interaction pattern for HDAC6 engagement, it is important to emphasize that isozyme selectivity was not explicitly assessed in our computational protocol and has not yet been validated experimentally. Therefore, further studies are required to evaluate the selectivity profile and biological activity of these compounds across different HDAC isozymes.
After confirming the ability of Compound 10 to enzymatically inhibit HDAC6, we exploited MM as a model to investigate its anti-tumor activity. It is well established that HDAC6 plays a critical role in MM pathogenesis by regulating protein homeostasis and aggresome formation, thereby enabling MM cells to survive proteotoxic stress, particularly under proteasome inhibition [ 55 ]. HDAC6 inhibitors have demonstrated promising anti-MM activity by disrupting aggresome-mediated protein degradation and enhancing the cytotoxic effects of PIs, thereby offering a potential strategy to overcome drug resistance in MM [ 20 , 56 ], supporting the therapeutic relevance of HDAC6 targeting in this cancer.
Of note, Compound 10 was able to inhibit MM cell viability in a mid-high micromolar range, with a lower IC 50 observed in PI-sensitive cells. This differential sensitivity may reflect distinct vulnerabilities in the HDAC6 pathway between these subtypes and supports the potential use of Compound 10 either as a monotherapy or in combination with PIs to overcome drug resistance. These findings underscore the therapeutic potential of Compound 10 in MM and warrant further mechanistic and combinatorial studies to evaluate its efficacy in overcoming resistance to existing therapies.
Supporting its potential on-target mechanism of action, Compound 10 treatment increased acetylation of the HDAC6 substrate α-tubulin in treated MM cell lines, while effects on acetylated histone H4 was significantly less pronounced.
Although these findings are promising, the selectivity profile of Compound 10 against other HDAC isoforms and unrelated enzymes remains to be fully established. We acknowledge this as a limitation of the current study and consider it an important objective for follow up investigations.
However, taken together, our data demonstrate that the anti-tumor effects of Compound 10 are closely associated with its ability to inhibit HDAC6 enzymatic activity, resulting in enhanced acetylation of downstream substrates. Finally, we can highlight the therapeutic relevance of targeting HDAC6 in MM and propose Compound 10 as a promising candidate for further development.
4. Materials and Methods
4.1. Pharmacophore Model Generation
A set of thirty HDAC6 inhibitors ( Table S1 ) was selected for ligand-based pharmacophore modeling based on their potent antitumor activity, defined by IC 50 values below 100 nM and their demonstrated selectivity toward HDAC6 across various cancer models. Particular emphasis was placed on hydroxamate-based structures, given their prevalence in publicly available datasets and their well-characterized zinc-binding modes within HDAC active sites. While the hydroxamic acid moiety is a common pharmacophoric feature across HDAC inhibitors, HDAC6 selectivity is primarily influenced by the structural variability in the surface recognition domain or capping group (CAP). Therefore, the training set was carefully curated to include a diverse array of CAP structures, ensuring that this key element of selectivity was well-represented in the pharmacophore hypothesis. Ligand-based pharmacophore modeling was carried out using this training set, and the resulting hypothesis was validated with a test set comprising both active and inactive compounds to evaluate its predictive accuracy and discriminative power [ 43 , 57 ].
Initially, the set of thirty compounds was clustered based on chemical similarity using the Tanimoto coefficient. The optimal number of clusters was determined with the Canvas tool [ 44 ], employing the Kelly criterion as the selection strategy, which resulted in six distinct groups. The centroid molecule from each cluster, representing six structurally diverse scaffolds, was selected as the ‘training set’ for pharmacophore model generation ( Table 1 ). These representative compounds were then used to derive a shared pharmacophore hypothesis within the Ligand-Based Modeling Perspective of LigandScout (version 4.4.6) [ 43 ]. All 2D molecular structures were initially constructed using the Maestro graphical interface of the Schrödinger Suite [ 58 ] and subsequently converted to the LBD format via the “Idbgen” utility in LigandScout. Conformer generation was performed using the OMEGA software v 3.1.1.2 (best settings), allowing for up to 200 conformations per molecule. LigandScout then derived common pharmacophore features by aligning the conformers and interpolating shared interactions based on a predefined number of optimal alignment solutions. A total of ten pharmacophore hypotheses were generated within a feature tolerance threshold of 1.0 Å and ranked using combined scoring metrics based on pharmacophore fit and atom overlap ( Table S2 ).
Accordingly, the generated pharmacophore model was retrospectively validated to assess its effectiveness in a virtual screening (VS) context. For this purpose, two distinct libraries comprising active, inactive compounds and decoys were assembled ( Tables S3–S5 ). Specifically, 12 known HDAC6 active compounds ( Table S3 ) were used to generate 375 property-matched decoys (at a 1:58 ratio) via the DUD-E database ( Table S4 ) [ 59 ]. These were combined with an additional set of 316 inactive compounds ( Table S5 ) retrieved from ChEMBL [ 60 ], resulting in a validation set of 703 molecules ( Tables S2–S4 ). The virtual screening of this dataset, followed by receiver operating characteristic (ROC) curve analysis, was employed to evaluate the model’s discriminatory capability.
4.2. Protein and Database Preparation and Docking Simulation
In our SBVS study, we carefully selected and downloaded the 3D coordinates of HDAC isozyme 6 from the Protein Data Bank (PDB) [ 61 ]. In particular, we used the crystallographic structures with the following PDB code: 5EDU [ 40 ]. We selected the protein structure based on three factors: (1) high resolution, (2) absence of mutations and (3) presence of co-crystallized inhibitor, as it provides valuable information about ligand binding and protein’s function.
The structures were prepared and energy-optimized through the Protein Preparation Wizard [ 62 ] tool implemented in Maestro [ 58 ], using OPLS_2005 as force field [ 63 ]. Thus, residual crystallographic buffer components and water molecules were removed, hydrogen atoms were added, and side chains’ protonation states were assigned at pH 7.4. The protein used consists of two chains, chain A and chain B, both complexed with Trichostatin A .
The docking grid was defined automatically based on the binding site of the co-crystallized ligand, Trichostatin A , which is present in both protein chains. Glide SP docking was performed using default settings to generate ten poses per ligand [ 64 ]. Re-docking of the reference ligand resulted in a lower root-mean-square deviation for chain A (1.12 Å) compared to chain B (5.67 Å), indicating more accurate pose reproduction and thus supporting the selection of chain A for downstream virtual screening ( Figure S10 ). To filter potential hits , the Glide Score (G-Score) of the best redocked pose of Trichostatin A (–5.52 kcal/mol) was used as the threshold.
4.3. Virtual Screening Strategy and Workflow
The virtual screening was conducted against commercial compound libraries from Asinex [ 42 ] and MolPort [ 41 ], collectively comprising over 300,000 molecules, including the Asinex BioDesign and Elite Libraries. The libraries were initially prepared using the LigPrep tool [ 65 ], where hydrogens were added, salts removed, and ionization states predicted at pH 7.4 using Epik. Each structure was energy-minimized using OPLS_2005 as a force field [ 63 ]. To avoid stereochemical ambiguities and complications from enantiomers, all chiral molecules were excluded, retaining only achiral compounds for screening.
Ligand-Based Virtual Screening (LBVS) was performed using LigandScout’s Screening Perspective with a validated pharmacophore model to identify compounds matching the predefined features. The resulting hits were further docked using the Glide SP protocol with default settings, generating up to ten poses per ligand. Compounds with Glide G-Scores better than the established HDAC6 threshold were selected for subsequent analysis. Fingerprint clustering based on the Tanimoto coefficient was then performed to group structurally similar compounds. Using the Kelley criterion, 4 optimal clusters were identified [ 44 ]. From each cluster, the molecule closest to the centroid—exhibiting the most shared features—was chosen as the representative hit .
4.4. Molecular Dynamic Simulations and MM-GBSA Analysis
The complexes of the four selected molecules with HDAC6 were subjected to 200 ns molecular dynamic simulations using Desmond version 4.4 [ 66 ], employing OPLS_2005 as a force field [ 63 ]. As a positive control, Trichostatin A complexed with the Class IIb HDAC6 isozyme was also simulated. Each system was placed within an orthorhombic simulation box with a 10 Å buffer and solvated explicitly using TIP4P as a water model [ 67 ]. Counterions were added to neutralize the overall system charge.
The reversible reference system propagation algorithm (r-RESPA) was applied with multiple time steps: 2 fs for bonded and short-range non-bonded interactions and 6 fs for long-range non-bonded interactions. For short-range Coulombic interactions, a time step of 1 fs and a 9.0 Å cutoff were used. Long-range electrostatics were treated with the Smooth Particle Mesh Ewald (PME) method [ 68 ].
Following solvation, energy minimization and equilibration were carried out using the Martyna–Tobias–Klein (MTK_NPT) ensemble. The systems were equilibrated under NVT and NPT ensembles at 1 atm pressure and 300 K temperature, controlled via the Berendsen thermostat and barostat. Trajectory frames were recorded every 20 ps and analyzed using the Simulation Interaction Diagram (SID) tool to assess geometric and thermodynamic properties of the complexes.
Finally, binding free energies (ΔG bind ) were estimated through Molecular Mechanics–Generalized Born Surface Area calculations [ 69 ], performed every 10 frames to evaluate the thermodynamic stability of the complexes.
4.5. Cell Cultures and Drugs
The MM cell lines used in this study were: AMO, AMO-BZB, NCI-H929 and NCI-H929-BZB. AMO and AMO-BZB cells were provided by Dr. C. Driessen (University of Tubingen, Tübingen, Germany). The NCI-H929 cell line was purchased from DSMZ (Braunschweig, Germany). Bortezomib (BZB)-resistant NCI-H929 cells were selected by incubation with increasing concentrations of the drug. Human MM cell lines were cultured in RPMI-1640 medium (Corning, Turin, Italy) supplemented with 10% of heat-inactivated fetal bovine serum (FBS) (Gibco ® , Life Technologies, Carlsbad, CA, USA), 100 U/mL of Penicillin and 100 µg/mL of Streptomycin (P/S) (Gibco ® , Life Technologies, Carlsbad, CA, USA) and incubated at 37 °C in a 5% CO 2 atmosphere. Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-hypaque (Lonza Group, Basel, Switzerland) from healthy donors, following informed consent and Institutional Review Board (University of Catanzaro, Catanzaro, Italy) approval (institutional approval: n.266/2021). PBMCs were cultured in RPMI-1640 medium (Gibco ® , Life Technologies) supplemented with 20% fetal bovine serum (Lonza Group Ltd., Basel, Switzerland) and 1% penicillin/streptomycin (Gibco ® , Life Technologies), as reported [ 70 ]. The 293T cell line was purchased from DSMZ and cultured in DMEM (Gibco ® , Life Technologies) supplemented with 10% fetal bovine serum (Lonza Group Ltd.) and 1% penicillin/streptomycin (Gibco ® , Life Technologies).
The selective HDAC6 inhibitor ACY-1215 (HY-16026) was purchased from MedChemExpress (MCE, Monmouth Junction, NJ, USA).
4.6. Cell Viability and Proliferation Assay
Cell viability was assessed using the CellTiter-Glo (CTG) assay kit (Promega, Madison, WI, USA), as reported [ 32 ]. Briefly, cells were seeded in 24-well plates and treated with Compound 10 at various concentrations. Luminescence was measured using the GloMax-Multi Detection System (Promega, Madison, WI, USA).
4.7. HDAC6 Inhibition In Vitro Assay
HDAC6 enzymatic activity was assessed using a fluorogenic HDAC6 assay kit (BPS Bioscience, cat#50076, San Diego, CA, USA), which includes a fluorometric substrate containing an acetylated lysine side chain. Upon deacetylation by HDAC6, the substrate becomes reactive to the Lysine Developer, which generates a fluorophore detectable using a fluorescence reader. Compounds AKOS030273637 ( 10 ), AKOS030461429 ( 9 ), AKOS000531201 ( 8 ), ACY-1215 and Trichostatin A were diluted in HDAC assay buffer and mixed with HDAC substrate (200 µM), BSA (1 mg/mL), HDAC assay buffer and HDAC6 human recombinant enzyme (14 ng/µL). The reaction mixtures were incubated at 37 °C for 30 min, and then the HDAC developer was added, and the mixtures were incubated at room temperature for 15 min. Fluorescence was measured using the GloMax-Multi Detection System (Promega, Madison, WI, USA). The dose–response curve and IC 50 values were determined using GraphPad Prism based on increasing concentrations (0.1, 0.25, 0.5, 1.0, 2.5, 5.0, 10.0, 25.0 and 50.0 mM) of Compound 10 .
4.8. Western Blotting (WB) Assays
Total protein extracts were prepared using NP-40 Cell Lysis Buffer supplemented with Halt™ Protease Inhibitor Cocktail (Thermo Fisher Scientific, Waltham, MA, USA), as reported [ 70 ].
Protein samples (30 µg per lane) were resolved on 10% SDS-PAGE gels and subsequently transferred onto nitrocellulose membranes using the Trans-Blot Turbo Transfer System (Bio-Rad Laboratories, Hercules, CA, USA). Membranes were then blocked and incubated with the following primary antibodies: acetylated-α Tubulin (#sc-23950; Santa Cruz Biotechnology, Dallas, TX, USA), acetylated-histone H4 (#sc-34263; Santa Cruz Biotechnology) or histone H4 (#AB5823; abcam). GAPDH (MA5-15738; Thermo Fisher Scientific, Waltham, MA, USA) was used as loading control. Protein bands were visualized using the Clarity Western ECL Substrate (Bio-Rad Laboratories, Hercules, CA, USA), and chemiluminescence signals were quantified using the UVITEC Imaging System (Cleaver Scientific, Warwickshire, UK).
5. Conclusions
Through an integrated computational and experimental approach, Compound 10 was identified as a promising HDAC6 inhibitor with favorable drug-like properties and stable binding behavior. It showed strong inhibitory in vitro activity and demonstrated anti-tumor effects in MM models, particularly in proteasome inhibitor-sensitive cells. The compound increased α-tubulin acetylation, supporting its on-target action. These findings suggest that Compound 10 represents a promising lead for further development in MM therapy. However, additional studies are required to assess its isoform selectivity profile, in vivo efficacy, and potential predictive biomarkers of response. Moreover, a comprehensive evaluation of its pharmacokinetic, pharmacodynamic, and toxicity profiles will be essential to support its progression toward preclinical development. Future work will focus on these aspects to better define its therapeutic potential.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Supplementary Materials
The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/ph18091303/s1 , Table S1. List of 31 active compounds used as the training set for ligand-based pharmacophore modeling; Table S2. Pharmacophores; Figure S1. Pharmacophore models mapped onto the training set compounds; Figure S2. Clustering analysis; Figure S3. RMSD Analysis of HDAC6–Ligand Complexes; Figure S4. Protein–ligand interactions between Trichostatin A and HDAC6 residues; Table S3. List of active compounds used for pharmacophore validation; Table S4. List of decoy compounds generated from the DUD-E database; Table S5. List of inactive compounds retrieved from ChEMBL; Figure S5. Dose–response curve analysis of HDAC6 activity in the presence of increasing concentrations of Compound 10. Figure S6 . Heatmap showing cell viability assessed by the CellTiter-Glo assay in MM cell lines treated with ACY-1215 or vehicle control (DMSO) for 72 h; Figure S7. Cell viability assessed by the CellTiter-Glo assay in 293T (A) or healthy PBMCs (B), 72 h after treatment with Compound 10 ; Figure S8. Western blot analysis of acetylated α-tubulin in AMO and AMO-BZB cells, 48 h after treatment with ACY-1215 ; Figure S9. Western Blot Analysis of Histone H4 and α-Tubulin Acetylation in AMO Cells Treated with Compound 10 and ACY1215; Figure S10. Three-dimensional visualization of redocking analysis.
Author Contributions
Conceptualization, A.A., R.R. and N.A.; methodology, F.C., A.C. and I.V.; software, S.A.; validation, F.C., A.C. and I.V.; formal analysis, R.R. and N.A.; investigation, F.C. and A.C.; resources, M.G. and A.A.; data curation, F.C. and A.C.; writing—original draft preparation, R.R. and N.A.; writing—review and editing, A.A. and S.A.; visualization, all authors; supervision, A.A.; project administration, A.A.; funding acquisition, S.A. and N.A. All authors have read and agreed to the published version of the manuscript.
Institutional Review Board Statement
Not applicable.
Informed Consent Statement
Not applicable.
Data Availability Statement
Data is contained within the article or Supplementary Materials .
Conflicts of Interest
The authors declare no conflicts of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.
Abbreviations
The following abbreviations are used in this manuscript: AUC Area under the curve BSA Bovine Serum Albumin BZB Bortezomib CD1 Catalytic domain 1 CD2 Catalytic domain 2 DMSO Dimethyl Sulfoxide DNA Deoxyribonucleic Acid DUD-E Directory of Useful Decoys, Enhanced FBS Fetal Bovine Serum FDA Food and Drug Administration GAPDH Glyceraldehyde 3-Phosphate Dehydrogenase HBA Hydrogen bond acceptor HBD Hydrogen bond donor HDAC Histone Deacetylase HDACis Histone Deacetylase Inhibitors HY Ionizable group IC 50 Half maximal inhibitory concentration IUPAC International Union of Pure and Applied Chemistry K Kelvin LBPS Ligand-Based Pharmacophore Screening LBVS Ligand-Based Virtual Screening MCE MedChemExpress MDs Molecular Dynamic simulation MM Multiple Myeloma MM-GBSA Molecular Mechanics Generalized Born Surface Area MTK_NPT Martyna-Tobias-Klein for NPT ensemble μM Micromolar NF-kB Nuclear Factor kappa-light-chain-enhancer of activated B cells NI Hydrophobic feature ns Nanosecond OPLS Optimized Potentials for Liquid Simulations PDB Protein Data Bank PIs Proteasome Inhibitors PME Particle Mesh Ewald r-RESPA Reversible Reference System Propagation Algorithm RMSD Root Mean Square Deviation ROC Receiver Operating Characteristic SBVS Structure-Based Virtual Screening SID Simulation Interaction Diagram SP Standard Precision TIP4P Transferable Intermolecular Potential with 4 Points TSA Trichostatin A VS Virtual Screening ZBG Zinc-binding group ZnF UBP Zinc finger ubiquitin-binding domain
References
1. Ho T.C.S. Chan A.H.Y. Ganesan A. Thirty Years of HDAC Inhibitors: 2020 Insight and Hindsight J. Med. Chem. 2020 63 12460 12484 10.1021/acs.jmedchem.0c00830 32608981
2. McClure J.J. Inks E.S. Zhang C. Peterson Y.K. Li J. Chundru K. Lee B. Buchanan A. Miao S. Chou C.J. Comparison of the Deacylase and Deacetylase Activity of Zinc-Dependent HDACs ACS Chem. Biol. 2017 12 1644 1655 10.1021/acschembio.7b00321 28459537 PMC5567816
3. Grozinger C.M. Hassig C.A. Schreiber S.L. Three proteins define a class of human histone deacetylases related to yeast Hda1p Proc. Natl. Acad. Sci. USA 1999 96 4868 4873 10.1073/pnas.96.9.4868 10220385 PMC21783
4. Ouyang H. Ali Y.O. Ravichandran M. Dong A. Qiu W. MacKenzie F. Dhe-Paganon S. Arrowsmith C.H. Zhai R.G. Protein aggregates are recruited to aggresome by histone deacetylase 6 via unanchored ubiquitin C termini J. Biol. Chem. 2012 287 2317 2327 10.1074/jbc.M111.273730 22069321 PMC3268394
5. Miyake Y. Keusch J.J. Wang L. Saito M. Hess D. Wang X. Melancon B.J. Helquist P. Gut H. Matthias P. Structural insights into HDAC6 tubulin deacetylation and its selective inhibition Nat. Chem. Biol. 2016 12 748 754 10.1038/nchembio.2140 27454931
6. Porter N.J. Mahendran A. Breslow R. Christianson D.W. Unusual zinc-binding mode of HDAC6-selective hydroxamate inhibitors Proc. Natl. Acad. Sci. USA 2017 114 13459 13464 10.1073/pnas.1718823114 29203661 PMC5754815
7. Osko J.D. Christianson D.W. Structural determinants of affinity and selectivity in the binding of inhibitors to histone deacetylase 6 Bioorg. Med. Chem. Lett. 2020 30 127023 10.1016/j.bmcl.2020.127023 32067866 PMC7067655
8. Porter N.J. Wagner F.F. Christianson D.W. Entropy as a Driver of Selectivity for Inhibitor Binding to Histone Deacetylase 6 Biochemistry 2018 57 3916 3924 10.1021/acs.biochem.8b00367 29775292 PMC6030506
9. Pulya S. Amin S.A. Adhikari N. Biswas S. Jha T. Ghosh B. HDAC6 as privileged target in drug discovery: A perspective Pharmacol. Res. 2021 163 105274 10.1016/j.phrs.2020.105274 33171304
10. Kawaguchi Y. Kovacs J.J. McLaurin A. Vance J.M. Ito A. Yao T.P. The deacetylase HDAC6 regulates aggresome formation and cell viability in response to misfolded protein stress Cell 2003 115 727 738 10.1016/S0092-8674(03)00939-5 14675537
11. Kanno K. Kanno S. Nitta H. Uesugi N. Sugai T. Masuda T. Wakabayashi G. Maesawa C. Overexpression of histone deacetylase 6 contributes to accelerated migration and invasion activity of hepatocellular carcinoma cells Oncol. Rep. 2012 28 867 873 10.3892/or.2012.1898 22766642
12. Wang Z. Hu P. Tang F. Lian H. Chen X. Zhang Y. He X. Liu W. Xie C. HDAC6 promotes cell proliferation and confers resistance to temozolomide in glioblastoma Cancer Lett. 2016 379 134 142 10.1016/j.canlet.2016.06.001 27267806
13. Witt O. Deubzer H.E. Milde T. Oehme I. HDAC family: What are the cancer relevant targets? Cancer Lett. 2009 277 8 21 10.1016/j.canlet.2008.08.016 18824292
14. Aldana-Masangkay G.I. Sakamoto K.M. The role of HDAC6 in cancer J. Biomed. Biotechnol. 2011 2011 875824 10.1155/2011/875824 21076528 PMC2975074
15. Reardon D.A. Egorin M.J. Quinn J.A. Rich J.N. Gururangan S. Vredenburgh J.J. Desjardins A. Sathornsumetee S. Provenzale J.M. Herndon J.E. Phase II study of imatinib mesylate plus hydroxyurea in adults with recurrent glioblastoma multiforme J. Clin. Oncol. 2005 23 9359 9368 10.1200/JCO.2005.03.2185 16361636
16. Slingerland M. Guchelaar H.J. Gelderblom H. Histone deacetylase inhibitors: An overview of the clinical studies in solid tumors Anticancer Drugs 2014 25 140 149 10.1097/CAD.0000000000000040 24185382
17. Gu X. Zhang H. Jiao M. Han B. Zhang Z. Li J. Zhang Q. Histone deacetylase 6 inhibitors with blood-brain barrier penetration as a potential strategy for CNS-Disorders therapy Eur. J. Med. Chem. 2022 229 114090 10.1016/j.ejmech.2021.114090 34992037
18. Lv W. Zhang G. Barinka C. Eubanks J.H. Kozikowski A.P. Design and Synthesis of Mercaptoacetamides as Potent, Selective, and Brain Permeable Histone Deacetylase 6 Inhibitors ACS Med. Chem. Lett. 2017 8 510 515 10.1021/acsmedchemlett.7b00012 28523102 PMC5430401
19. El Omari N. Bakrim S. Elhrech H. Aanniz T. Balahbib A. Lee L.H. Al Abdulmonem W. Bouyahya A. Clinical efficacy and mechanistic insights of FDA-approved HDAC inhibitors in the treatment of lymphoma Eur. J. Pharm. Sci. 2025 208 107057 10.1016/j.ejps.2025.107057 40043823
20. Santo L. Hideshima T. Kung A.L. Tseng J.C. Tamang D. Yang M. Jarpe M. van Duzer J.H. Mazitschek R. Ogier W.C. Preclinical activity, pharmacodynamic, and pharmacokinetic properties of a selective HDAC6 inhibitor, ACY-1215, in combination with bortezomib in multiple myeloma Blood 2012 119 2579 2589 10.1182/blood-2011-10-387365 22262760 PMC3337713
21. Bergman J.A. Woan K. Perez-Villarroel P. Villagra A. Sotomayor E.M. Kozikowski A.P. Selective histone deacetylase 6 inhibitors bearing substituted urea linkers inhibit melanoma cell growth J. Med. Chem. 2012 55 9891 9899 10.1021/jm301098e 23009203 PMC3562128
22. Liu J.R. Yu C.W. Hung P.Y. Hsin L.W. Chern J.W. High-selective HDAC6 inhibitor promotes HDAC6 degradation following autophagy modulation and enhanced antitumor immunity in glioblastoma Biochem. Pharmacol. 2019 163 458 471 10.1016/j.bcp.2019.03.023 30885763
23. Peng J. Xie F. Qin P. Liu Y. Niu H. Sun J. Xue H. Zhao Q. Liu J. Wu J. Recent development of selective inhibitors targeting the HDAC6 as anti-cancer drugs: Structure, function and design Bioorg. Chem. 2023 138 106622 10.1016/j.bioorg.2023.106622 37244230
24. Amodio N. D’Aquila P. Passarino G. Tassone P. Bellizzi D. Epigenetic modifications in multiple myeloma: Recent advances on the role of DNA and histone methylation Expert. Opin. Ther. Targets 2017 21 91 101 10.1080/14728222.2016.1266339 27892767
25. Paradzik T. Bandini C. Mereu E. Labrador M. Taiana E. Amodio N. Neri A. Piva R. The Landscape of Signaling Pathways and Proteasome Inhibitors Combinations in Multiple Myeloma Cancers 2021 13 6 10.3390/cancers13061235 33799793 PMC8000754
26. D’Aquila P. Ronchetti D. Gallo Cantafio M.E. Todoerti K. Taiana E. Fabiani F. Montesanto A. Neri A. Passarino G. Viglietto G. Epigenetic Regulation of Mitochondrial Quality Control Genes in Multiple Myeloma: A Sequenom MassARRAY Pilot Investigation on HMCLs J. Clin. Med. 2021 10 6 10.3390/jcm10061295 PMC8004002 33801014
27. Catalano R. Rocca R. Juli G. Costa G. Maruca A. Artese A. Caracciolo D. Tagliaferri P. Alcaro S. Tassone P. A drug repurposing screening reveals a novel epigenetic activity of hydroxychloroquine Eur. J. Med. Chem. 2019 183 111715 10.1016/j.ejmech.2019.111715 31550663
28. Gallo Cantafio M.E. Torcasio R. Viglietto G. Amodio N. Non-Coding RNA-Dependent Regulation of Mitochondrial Dynamics in Cancer Pathophysiology Noncoding RNA 2023 9 1 10.3390/ncrna9010016 36827549 PMC9964195
29. Gallo Cantafio M.E. Torcasio R. Scionti F. Mesuraca M. Ronchetti D. Pistoni M. Bellizzi D. Passarino G. Morelli E. Neri A. GPER1 Activation Exerts Anti-Tumor Activity in Multiple Myeloma Cells 2023 12 18 10.3390/cells12182226 PMC10526814 37759449
30. Torcasio R. Gallo Cantafio M.E. Ikeda R.K. Ganino L. Viglietto G. Amodio N. Lipid metabolic vulnerabilities of multiple myeloma Clin. Exp. Med. 2023 23 3373 3390 10.1007/s10238-023-01174-2 37639069 PMC10618328
31. Becherini P. Soncini D. Ravera S. Gelli E. Martinuzzi C. Giorgetti G. Cagnetta A. Guolo F. Ivaldi F. Miglino M. CD38-Induced Metabolic Dysfunction Primes Multiple Myeloma Cells for NAD Antioxidants 2023 12 2 10.3390/antiox12020494 36830052 PMC9952390
32. Torcasio R. Gallo Cantafio M.E. Veneziano C. De Marco C. Ganino L. Valentino I. Occhiuzzi M.A. Perrotta I.D. Mancuso T. Conforti F. Targeting of mitochondrial fission through natural flavanones elicits anti-myeloma activity J. Transl. Med. 2024 22 208 10.1186/s12967-024-05013-0 38413989 PMC10898065
33. Marchese E. Gallo Cantafio M.E. Ambrosio F.A. Torcasio R. Valentino I. Trapasso F. Viglietto G. Alcaro S. Costa G. Amodio N. New Insights for Polyphenolic Compounds as Naturally Inspired Proteasome Inhibitors Pharmaceuticals 2023 16 12 10.3390/ph16121712 38139838 PMC10747119
34. Ambrosio F.A. Costa G. Gallo Cantafio M.E. Torcasio R. Trapasso F. Alcaro S. Viglietto G. Amodio N. Natural Agents as Novel Potential Source of Proteasome Inhibitors with Anti-Tumor Activity: Focus on Multiple Myeloma Molecules 2023 28 3 10.3390/molecules28031438 PMC9919276 36771100
35. Todoerti K. Gallo Cantafio M.E. Oliverio M. Juli G. Rocca C. Citraro R. Tassone P. Procopio A. De Sarro G. Neri A. Oleil Hydroxytyrosol (HTOL) Exerts Anti-Myeloma Activity by Antagonizing Key Survival Pathways in Malignant Plasma Cells Int. J. Mol. Sci. 2021 22 21 10.3390/ijms222111639 PMC8584245 34769070
36. Ling H. Hongwei W. Zhiqiang S. Jie Z. Inhibition of Multiple Myeloma Growth by Wogonin Involves Mitochondrial Apoptosis and G2/M Cycle Arrest Int. J. Pharmacol. 2024 20 166 173 10.3923/ijp.2024.166.173
37. Gao X.-H. Liu Y.-J. Liu Y.-T. Wang Y.-Y. Yang J. Li Y. Li J. A Retrospective Analysis of Morbidity and Risk Factors of Multiple Myeloma with Peripheral Neuropathy Br. J. Hosp. Med. 2024 85 1 10 10.12968/hmed.2024.0544 39831501
38. Liu F. Liu C. Chai Q. Zhao C. Meng H. Xue X. Yao T.P. Zhang Y. Discovery of the First Irreversible HDAC6 Isoform Selective Inhibitor with Potent Anti-Multiple Myeloma Activity J. Med. Chem. 2023 66 10080 10091 10.1021/acs.jmedchem.3c00977 37463038
39. Zhang X.H. Qin-Ma Wu H.P. Khamis M.Y. Li Y.H. Ma L.Y. Liu H.M. A Review of Progress in Histone Deacetylase 6 Inhibitors Research: Structural Specificity and Functional Diversity J. Med. Chem. 2021 64 1362 1391 10.1021/acs.jmedchem.0c01782 33523672
40. Hai Y. Christianson D.W. Histone deacetylase 6 structure and molecular basis of catalysis and inhibition Nat. Chem. Biol. 2016 12 741 747 10.1038/nchembio.2134 27454933 PMC4990478
41. Sorokina M. Merseburger P. Rajan K. Yirik M.A. Steinbeck C. COCONUT online: Collection of Open Natural Products database J. Cheminform. 2021 13 2 10.1186/s13321-020-00478-9 33423696 PMC7798278
42. Asinex Available online: http://www.asinex.com/ (accessed on 15 June 2024)
43. Wolber G. Langer T. LigandScout: 3-D pharmacophores derived from protein-bound ligands and their use as virtual screening filters J. Chem. Inf. Model. 2005 45 160 169 10.1021/ci049885e 15667141
44. Schrödinger Canvas Schrödinger, LLC New York, NY, USA 2020
45. Yan G. Li D. Zhong X. Liu G. Wang X. Lu Y. Qin F. Guo Y. Duan S. Identification of HDAC6 selective inhibitors: Pharmacophore based virtual screening, molecular docking and molecular dynamics simulation J. Biomol. Struct. Dyn. 2021 39 1928 1939 10.1080/07391102.2020.1743760 32178584
46. Song H. Niu X. Quan J. Li Y. Yuan L. Wang J. Ma C. Ma E. Discovery of specific HDAC6 inhibitor with anti-metastatic effects in pancreatic cancer cells through virtual screening and biological evaluation Bioorganic Chem. 2020 97 103679 10.1016/j.bioorg.2020.103679 32120077
47. Liu X. Yan W. Wang S. Lu M. Yang H. Chai X. Shi H. Zhang Y. Jia Q. Discovery of selective HDAC6 inhibitors based on a multi-layer virtual screening strategy Comput. Biol. Med. 2023 160 107036 10.1016/j.compbiomed.2023.107036 37196455
48. Goracci L. Deschamps N. Randazzo G.M. Petit C. Dos Santos Passos C. Carrupt P.A. Simões-Pires C. Nurisso A. A Rational Approach for the Identification of Non-Hydroxamate HDAC6-Selective Inhibitors Sci. Rep. 2016 6 29086 10.1038/srep29086 27404291 PMC4941420
49. Curcio A. Rocca R. Alcaro S. Artese A. The Histone Deacetylase Family: Structural Features and Application of Combined Computational Methods Pharmaceuticals 2024 17 5 10.3390/ph17050620 PMC11124352 38794190
50. Wang X.X. Wan R.Z. Liu Z.P. Recent advances in the discovery of potent and selective HDAC6 inhibitors Eur. J. Med. Chem. 2018 143 1406 1418 10.1016/j.ejmech.2017.10.040 29133060
51. Daina A. Michielin O. Zoete V. SwissADME: A free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules Sci. Rep. 2017 7 42717 10.1038/srep42717 28256516 PMC5335600
52. Rossi C. Porcelloni M. D’Andrea P. Fincham C.I. Ettorre A. Mauro S. Squarcia A. Bigioni M. Parlani M. Nardelli F. Alkyl piperidine and piperazine hydroxamic acids as HDAC inhibitors Bioorg. Med. Chem. Lett. 2011 21 2305 2308 10.1016/j.bmcl.2011.02.085 21420859
53. Rossi C. Porcelloni M. D’Andrea P. Fattori D. Marastoni E. Hydroxamates as His-Tone Deacetylase Inhibitors and Pharmaceutical Formulations Containing Them WO2006097460 14 March 2006
54. Olaoye O.O. Watson P.R. Nawar N. Geletu M. Sedighi A. Bukhari S. Raouf Y.S. Manaswiyoungkul P. Erdogan F. Abdeldayem A. Unique Molecular Interaction with the Histone Deacetylase 6 Catalytic Tunnel: Crystallographic and Biological Characterization of a Model Chemotype J. Med. Chem. 2021 64 2691 2704 10.1021/acs.jmedchem.0c01922 33576627 PMC8063965
55. Lee S.W. Yeon S.K. Kim G.W. Lee D.H. Jeon Y.H. Yoo J. Kim S.Y. Kwon S.H. HDAC6-Selective Inhibitor Overcomes Bortezomib Resistance in Multiple Myeloma Int. J. Mol. Sci. 2021 22 3 10.3390/ijms22031341 33572814 PMC7866276
56. Mishima Y. Santo L. Eda H. Cirstea D. Nemani N. Yee A.J. O’Donnell E. Selig M.K. Quayle S.N. Arastu-Kapur S. Ricolinostat (ACY-1215) induced inhibition of aggresome formation accelerates carfilzomib-induced multiple myeloma cell death Br. J. Haematol. 2015 169 423 434 10.1111/bjh.13315 25709080
57. Meslamani J. Li J. Sutter J. Stevens A. Bertrand H.O. Rognan D. Protein-ligand-based pharmacophores: Generation and utility assessment in computational ligand profiling J. Chem. Inf. Model. 2012 52 943 955 10.1021/ci300083r 22480372
58. Schrödinger Maestro Schrödinger, LLC New York, NY, USA 2020
59. Mysinger M.M. Carchia M. Irwin J.J. Shoichet B.K. Directory of useful decoys, enhanced (DUD-E): Better ligands and decoys for better benchmarking J. Med. Chem. 2012 55 6582 6594 10.1021/jm300687e 22716043 PMC3405771
60. Mendez D. Gaulton A. Bento A.P. Chambers J. De Veij M. Félix E. Magariños M.P. Mosquera J.F. Mutowo P. Nowotka M. ChEMBL: Towards direct deposition of bioassay data Nucleic Acids Res. 2019 47 D930 D940 10.1093/nar/gky1075 30398643 PMC6323927
61. Berman H.M. Westbrook J. Feng Z. Gilliland G. Bhat T.N. Weissig H. Shindyalov I.N. Bourne P.E. The Protein Data Bank Nucleic Acids Res. 2000 28 235 242 10.1093/nar/28.1.235 10592235 PMC102472
62. Schrödinger Protein Preparation Wizard Schrödinger, LLC New York, NY, USA 2020
63. Shivakumar D. Harder E. Damm W. Friesner R.A. Sherman W. Improving the Prediction of Absolute Solvation Free Energies Using the Next Generation OPLS Force Field J. Chem. Theory Comput. 2012 8 2553 2558 10.1021/ct300203w 26592101
64. Schrödinger Glide Schrödinger, LLC New York, NY, USA 2020
65. Schrödinger LigPrep Schrödinger, LLC New York, NY, USA 2020
66. Schrödinger Desmond Schrödinger, LLC New York, NY, USA 2020
67. Jorgensen W.L. Chandrasekhar J. Madura J.D. Impey R.W. Klein M.L. Comparison of Simple Potential Functions for Simulating Liquid Water J. Chem. Phys. 1983 79 926 935 10.1063/1.445869
68. Darden T. York D. Pedersen L. Particle mesh Ewald: An Nlog (N) method for Ewald sums in large systems J. Chem. Phys. 1993 98 10089 10092 10.1063/1.464397
69. Schrödinger Prime Schrödinger, LLC New York, NY, USA 2020
70. Cantafio M.E.G. Valentino I. Torcasio R. Ganino L. Veneziano C. Murfone P. Mesuraca M. Perrotta I. Tallarigo F. Agosti V. Mitochondrial fission factor drives an actionable metabolic vulnerability in multiple myeloma Haematologica 2025 early view 10.3324/haematol.2025.287526 PMC12485323 40400476