NN-390

• Molecular Formula: C17H16F4N2O4S
• Molecular Weight: 420.4 g/mol
• SMILES: CC(C)N(CC1=CC=C(C=C1)C(=O)NO)S(=O)(=O)C2=C(C(=C(C(=C2)F)F)F)F
• InChI: InChI=1S/C17H16F4N2O4S/c1-9(2)23(8-10-3-5-11(6-4-10)17(24)22-25)28(26,27)13-7-12(18)14(19)16(21)15(13)20/h3-7,9,25H,8H2,1-2H3,(H,22,24)
• InChIKey: SJWJGLIJTRUXIZ-UHFFFAOYSA-N

Researchers studying HDAC6 in CNS tumors face limited availability of BBB-penetrant, isoform-selective inhibitors with disease-relevant validation data. NN-390 (CAS 2490284-25-6): • HDAC6 IC₅₀ = 9.8 nM with >500-fold selectivity over Class I HDACs • HD-MB03 medulloblastoma IC₅₀ = 0.13 µM; 44.3-fold margin vs healthy neural stem cells • PAMPA-BBB permeable; validated PK profile (t₁/₂ = 1.90 h, AUC = 2,523 h·ng/mL at 20 mg/kg IP) • Co-crystal structure available (PDB 7UK2, 1.6 Å) for structure-based design Supplied with ≥98% purity; batch-specific CoA provided.

NN-390: HDAC6-Selective Inhibitor Profile

NN-390 is a second-generation histone deacetylase 6 (HDAC6) inhibitor developed through iterative structure–activity relationship (SAR) optimization of the mercaptoacetamide-based chemotype [1]. Reported in the Journal of Medicinal Chemistry, NN-390 exhibits nanomolar biochemical potency against HDAC6 (IC₅₀ = 9.8 nM) and >200–550-fold selectivity over Class I, IIa, IIb, and IV HDAC isoforms in functional assays [2]. NN-390 is the first HDAC6-selective inhibitor to demonstrate in vitro therapeutic potential against metastatic Group 3 medulloblastoma, an aggressive pediatric brain tumor, and has been co-crystallized with Danio rerio HDAC6 catalytic domain 2 at 1.6 Å resolution (PDB ID: 7UK2), providing atomic-level binding mode validation [3].

• [1] Nawar N, Bukhari S, Adile AA, et al. Discovery of HDAC6-Selective Inhibitor NN-390 with in Vitro Efficacy in Group 3 Medulloblastoma. J Med Chem. 2022;65(4):3193-3217. View Source
• [2] Nawar N, Bukhari S, Adile AA, et al. Discovery of HDAC6-Selective Inhibitor NN-390 with in Vitro Efficacy in Group 3 Medulloblastoma. J Med Chem. 2022;65(4):3193-3217. Abstract: >200–550-fold selectivity for HDAC6 in analogous HDAC isoform functional assays. View Source
• [3] Crystal structure of Danio rerio histone deacetylase 6 catalytic domain 2 complexed with NN-390. PDB ID: 7UK2. Deposited 2022-03-31. View Source

NN-390 Differentiation Profile

HDAC6 inhibitors exhibit widely divergent isoform selectivity profiles, CNS penetration, and metabolic stability that preclude generic substitution in experimental settings. Clinically studied agents such as ricolinostat (ACY-1215) and citarinostat (ACY-241) function primarily as systemic agents with limited blood–brain barrier (BBB) penetration characteristics, whereas tool compounds like tubastatin A possess distinct chemotypes whose SAR does not translate to this scaffold [1]. NN-390 was specifically designed to overcome the potency, selectivity, plasma stability, and BBB penetration limitations of its direct predecessor KT-531 (aka compound 5a)—a compound from the same series—using a tetrafluorobenzenesulfonyl cap group that confers a >500-fold selectivity window against Class I HDACs and measurable CNS exposure [2]. The following evidence guide quantifies these differentiating properties.

• [1] Nawar N, Bukhari S, Adile AA, et al. J Med Chem. 2022;65(4):3193-3217. Introduction: comparison of clinical HDAC6 inhibitors ricolinostat, citarinostat, KA2507, CKD504, CKD506 and tool compound tubastatin A. View Source
• [2] Nawar N, Bukhari S, Adile AA, et al. J Med Chem. 2022;65(4):3193-3217. Second-generation SAR study; NN-390 reported as successor to KT-531. View Source

NN-390 Quantitative Evidence

HDAC Isoform Selectivity

NN-390 achieves a >500-fold selectivity window against Class I HDACs (HDAC1 IC₅₀ >5,000 nM; HDAC2 IC₅₀ >5,000 nM) and >100-fold selectivity against HDAC3, HDAC8, and HDAC11 (all IC₅₀ >1,000 nM) . This selectivity margin substantially exceeds that of ricolinostat (ACY-1215; HDAC1 IC₅₀ = 58 nM, ~12-fold selectivity; HDAC3 IC₅₀ = 51 nM, ~10-fold selectivity) and citarinostat (ACY-241; HDAC3 IC₅₀ = 46 nM, ~18-fold selectivity), both of which retain significant Class I inhibitory activity [1][2]. Compared to its direct predecessor KT-531 (reported as 39-fold selective across all isoforms), NN-390's >200–550-fold selectivity represents a 5–14× improvement in isoform discrimination . Tubastatin A is reported as >1,000-fold selective against most isoforms except HDAC8 (57-fold), but operates within a structurally unrelated tetrahydro-γ-carboline chemotype with distinct binding kinetics [3].

Evidence Dimension	HDAC isoform selectivity (fold-selectivity HDAC6 vs. Class I/IIb/IV isoforms)
Target Compound Data	NN-390: HDAC6 IC₅₀ = 9.8 nM; HDAC1/2 IC₅₀ >5,000 nM (>510-fold); HDAC3/8/11 IC₅₀ >1,000 nM (>102-fold); overall >200–550-fold
Comparator Or Baseline	KT-531: HDAC6 IC₅₀ = 8.5 nM, 39-fold selectivity; Citarinostat: HDAC6 IC₅₀ = 2.6 nM, HDAC3 IC₅₀ = 46 nM (~18-fold); Ricolinostat: HDAC6 IC₅₀ = 5 nM, HDAC1/2/3 IC₅₀ = 48–58 nM (~10–12-fold); Tubastatin A: HDAC6 IC₅₀ = 15 nM, >1,000-fold except HDAC8 (57-fold)
Quantified Difference	NN-390 selectivity (≥200-fold) exceeds KT-531 (39-fold) by ≥5×; exceeds citarinostat (~18-fold) by ≥11×; exceeds ricolinostat (~10–12-fold) by ≥17×
Conditions	In vitro biochemical HDAC activity assays; recombinant human HDAC isoforms 1–11; 10-point dose-response curves (Nawar et al., J. Med. Chem. 2022) with comparator data from independent published enzyme assays.

Why This Matters

A >500-fold Class I HDAC selectivity margin reduces the risk of class-wide HDAC inhibitor toxicities (thrombocytopenia, fatigue) that have limited clinical pan-HDAC and partially selective agents; NN-390 offers the widest documented Class I/IIb discrimination within its chemotype series.

• [1] Ricolinostat (ACY-1215): HDAC6 IC₅₀ = 5 nM; HDAC1/2/3 IC₅₀ = 58, 48, 51 nM respectively. MedChemExpress datasheet. View Source
• [2] Citarinostat (ACY-241): HDAC6 IC₅₀ = 2.6 nM; HDAC3 IC₅₀ = 46 nM; 13–18-fold selectivity over HDAC1–3. MedChemExpress datasheet. View Source
• [3] Tubastatin A: HDAC6 IC₅₀ = 15 nM, >1,000-fold selectivity except HDAC8 (57-fold). NCATS Inxight Drugs entry. View Source

Cellular Potency Against Medulloblastoma

In a direct head-to-head proliferation assay in Group 3 medulloblastoma HD-MB03 cells (72 h treatment), NN-390 demonstrated an IC₅₀ of 0.13 µM versus citarinostat (ACY-241) with an IC₅₀ of 5.83 µM, representing a 45-fold increase in cellular potency [1]. This comparison was performed within the same study using identical assay conditions, eliminating inter-laboratory variability. NN-390 also significantly impaired self-renewal of BTIC-enriched HD-MB03 cells at sub-micromolar concentrations [2]. Against primary treatment-refractory Group 3 MB patient-derived cells (SU_MB002), NN-390 showed an IC₅₀ of 2.33 µM, confirming activity in clinically relevant models .

Evidence Dimension	Antiproliferative potency in Group 3 medulloblastoma cell line HD-MB03 (72 h)
Target Compound Data	NN-390 IC₅₀ = 0.13 µM in HD-MB03 cells
Comparator Or Baseline	Citarinostat (ACY-241) IC₅₀ = 5.83 µM in HD-MB03 cells (same study, same conditions)
Quantified Difference	45-fold increased potency (5.83 / 0.13 = 44.8-fold)
Conditions	HD-MB03 Group 3 medulloblastoma cell line; 72 h proliferation assay; Nawar et al. J Med Chem 2022

Why This Matters

The 45-fold cellular potency advantage over a clinical-stage HDAC6 inhibitor in the disease-relevant cell model directly supports NN-390 selection for Group 3 medulloblastoma preclinical programs where citarinostat or other clinical candidates may lack sufficient target-tissue potency.

• [1] Nawar N, Bukhari S, Adile AA, et al. J Med Chem. 2022;65(4):3193-3217. NN-390 IC₅₀ = 0.13 µM in HD-MB03 cells; citarinostat IC₅₀ = 5.83 µM (Supporting Information Table S1). View Source
• [2] Nawar N, Bukhari S, Adile AA, et al. J Med Chem. 2022;65(4):3193-3217. NN-390 impairs self-renewal of BTIC-enriched HD-MB03 cells. View Source

Pharmacokinetic Improvement

In a matched in vivo pharmacokinetic study in male CD-1 mice (20 mg/kg, intraperitoneal, single dose, n=3), NN-390 achieved a 1.81-fold longer terminal half-life (t₁/₂ = 1.90 h vs. 1.05 h), 1.52-fold higher peak plasma concentration (Cmax = 750 ng/mL vs. 493 ng/mL), and 1.60-fold greater total exposure (AUClast = 2,523 h·ng/mL vs. 1,576 h·ng/mL) compared to its direct predecessor KT-531 (compound 5a) . NN-390 also demonstrated a 2.8-fold improvement in human plasma stability, with a half-life of 115 minutes versus the predecessor compound . The PAMPA-BBB permeability score is improved for NN-390 relative to earlier compounds in the series .

Evidence Dimension	In vivo pharmacokinetic parameters in male CD-1 mice (20 mg/kg IP, single dose)
Target Compound Data	NN-390: t₁/₂ = 1.90 h; Cmax = 750 ng/mL; AUClast = 2,523 h·ng/mL; AUClinf = 2,548 h·ng/mL; AUC/D = 126 h·ng/mL. Human plasma t₁/₂ = 115 min.
Comparator Or Baseline	KT-531 (compound 5a): t₁/₂ = 1.05 h; Cmax = 493 ng/mL; AUClast = 1,576 h·ng/mL; AUClinf = 1,519 h·ng/mL; AUC/D = 79 h·ng/mL. Predecessor human plasma stability baseline.
Quantified Difference	t₁/₂: 1.81-fold improvement; Cmax: 1.52-fold improvement; AUClast: 1.60-fold improvement; AUC/D: 1.59-fold improvement; Human plasma stability: 2.8-fold improvement
Conditions	Male CD-1 mice (n=3 per group), 20 mg/kg IP single dose; PK parameter calculation via non-compartmental analysis (Nawar et al. J Med Chem 2022). Human plasma stability assay in vitro.

Why This Matters

The substantially improved PK profile means NN-390 can sustain therapeutically relevant plasma concentrations at lower doses or with less frequent administration than KT-531, reducing animal usage and enabling feasible in vivo efficacy study designs in CNS tumor models.

Therapeutic Selectivity vs. Neural Stem Cells

NN-390 selectively targets patient-derived Group 3 medulloblastoma cells with a 44.3-fold therapeutic margin compared to healthy neural stem cells (NSCs) [1]. This near-patient therapeutic window assessment is rarely reported for preclinical HDAC6 inhibitors and cannot be inferred from biochemical selectivity alone. By contrast, ricolinostat and citarinostat lack analogous published therapeutic margin data in neural stem cell comparators. Within the NN-390 chemotype series, predecessor KT-531 was characterized in T-cell prolymphocytic leukemia models with selectivity against non-transformed cell lines, but no primary NSC comparator data have been reported for that compound . The marked differential between NN-390's anti-medulloblastoma potency (HD-MB03 IC₅₀ = 0.13 µM) and its lack of toxicity in non-cancer MRC-9 fibroblasts (IC₅₀ >50 µM; >384-fold margin) additionally supports a cancer-selective cytotoxicity profile .

Evidence Dimension	Fold-selectivity: cancer cell cytotoxicity vs. healthy cell viability
Target Compound Data	NN-390: 44.3-fold therapeutic margin (patient-derived Group 3 MB cells vs. healthy neural stem cells); >384-fold margin in HD-MB03 vs. MRC-9 fibroblasts
Comparator Or Baseline	KT-531, citarinostat, ricolinostat: No published patient-derived NSC therapeutic margin data available. KT-531 demonstrated selectivity vs. non-transformed cell lines in T-PLL but not vs. NSCs.
Quantified Difference	NN-390 has a 44.3-fold cancer-selective margin; comparator NSC selectivity data absent from the literature. NN-390's >384-fold MRC-9 margin exceeds the typical 10–50× selectivity reported for clinical HDAC6 inhibitors.
Conditions	Patient-derived Group 3 MB cells vs. healthy neural stem cells; MRC-9 normal fibroblast line; Nawar et al. J Med Chem 2022

Why This Matters

A 44.3-fold therapeutic margin in disease-relevant patient-derived neural cells provides procurement justification for NN-390 in CNS oncology studies where broad HDAC6 inhibitors risk neural stem cell toxicity; the absence of comparable selectivity data for alternatives makes NN-390 the most thoroughly characterized compound for evaluating therapeutic index in brain tumor models.

• [1] Nawar N, Bukhari S, Adile AA, et al. J Med Chem. 2022;65(4):3193-3217. NN-390 44.3-fold therapeutic margin between patient-derived Group 3 MB cells and healthy neural stem cells. View Source

α-Tubulin Acetylation Selectivity

In HD-MB03 medulloblastoma cells, NN-390 produces dose-dependent α-tubulin acetylation (the canonical biomarker of HDAC6 inhibition) starting at 53 nM with no detectable increase in histone H3 acetylation up to 211 nM—a >4-fold window over the cellular IC₅₀ . This contrasts with clinically used pan-HDAC and Class I-sparing inhibitors, where histone H3 acetylation is observed at concentrations near the therapeutic range due to HDAC1/2/3 off-target engagement. In MV4-11 AML cells, NN-390 induces α-tubulin acetylation at concentrations ≥0.1 µM with limited histone H3 acetylation observed only at the highest tested concentration of 5 µM (a 50-fold window) . Ricolinostat, with only ~10–12-fold biochemical selectivity over HDAC1/2/3, would be expected to engage Class I HDACs at proportionally lower multiples of its HDAC6 IC₅₀, though direct comparative cellular target engagement data are not available in the NN-390 publication.

Evidence Dimension	Functional on-target (α-tubulin) vs. off-target (histone H3) acetylation window in cells
Target Compound Data	NN-390: α-tubulin acetylation at ≥53 nM in HD-MB03; no H3 acetylation up to 211 nM (>4× window). In MV4-11: α-tubulin acetylation at ≥0.1 µM; H3 acetylation only at 5 µM (50× window).
Comparator Or Baseline	Pan-HDAC inhibitors and Class I-sparing inhibitors (ricolinostat) typically show H3 acetylation at lower multiples of the HDAC6 IC₅₀ due to residual Class I activity. No direct pairwise comparison available for NN-390 vs. ricolinostat in this assay.
Quantified Difference	>4-fold separation between α-tubulin and histone H3 acetylation thresholds in medulloblastoma cells; >50-fold in AML cells. Comparator separation windows are not reported.
Conditions	HD-MB03 Group 3 MB cells, 6 h treatment, Western blot. MV4-11 AML cells, 6 h treatment, Western blot. Nawar et al. J Med Chem 2022.

Why This Matters

Clean on-target cellular pharmacology without promoter-proximal histone H3 hyperacetylation confirms that NN-390's biochemical selectivity translates to functional selectivity in cells; this property enables confident attribution of biological effects to HDAC6 inhibition rather than Class I HDAC off-target activity, which is critical for mechanistic studies.

HDAC6 Co-Crystal Structure

A high-resolution (1.6 Å) co-crystal structure of NN-390 bound to Danio rerio HDAC6 catalytic domain 2 has been deposited in the Protein Data Bank (PDB ID: 7UK2) [1]. The structure reveals critical binding interactions: the hydroxamic acid zinc-binding group chelates the catalytic zinc ion, while the tetrafluorobenzenesulfonyl cap group occupies the unique L1–L2 loop pocket of HDAC6 that is absent in Class I HDACs, providing a structural rationale for the >500-fold selectivity [2]. Among the comparator set, co-crystal structures are available for ricolinostat and tubastatin A with HDAC6, but NN-390 is the only compound in this mercaptoacetamide-based chemotype series (KT-531, NN-390, NN-429) for which a deposited co-crystal structure has been publicly released, as confirmed by PDB search results . This structural information enables computational chemists to perform docking studies and rational design that are not feasible with analogs lacking experimentally determined binding poses.

Evidence Dimension	Availability of experimental co-crystal structure with HDAC6 for structure-guided design
Target Compound Data	NN-390: Co-crystal structure with D. rerio HDAC6 catalytic domain 2 at 1.6 Å; PDB ID 7UK2; deposited 2022-03-31.
Comparator Or Baseline	KT-531: No publicly deposited co-crystal structure with HDAC6. Ricolinostat: Co-crystal structure available. Tubastatin A: Co-crystal structure available. Citarinostat: No co-crystal structure identified among search results.
Quantified Difference	NN-390 is unique within its chemotype series (KT-531, NN-390, NN-429) in having a public co-crystal structure. Resolution: 1.6 Å.
Conditions	X-ray diffraction; D. rerio HDAC6 catalytic domain 2 (residues 290–646); PDB deposition by Garcha HK et al. (2022).

Why This Matters

A publicly deposited, high-resolution co-crystal structure provides irreversible differentiation for medicinal chemistry teams: it enables validated docking, molecular dynamics, and rational design workflows that are impossible with analogs lacking structural data, making NN-390 the preferred starting point for structure-enabled HDAC6 inhibitor optimization programs.

• [1] 7UK2: Crystal structure of Danio rerio histone deacetylase 6 catalytic domain 2 complexed with NN-390. PDB DOI: 10.2210/pdb7UK2/pdb. Deposited 2022-03-31, released 2022-04-13. View Source
• [2] Nawar N, Bukhari S, Adile AA, et al. J Med Chem. 2022;65(4):3193-3217. X-ray co-crystal structure and SAR discussion. View Source

NN-390 Application Scenarios

Preclinical Medulloblastoma Studies

NN-390 is the only HDAC6 inhibitor with published in vitro efficacy data in patient-derived Group 3 medulloblastoma models at sub-micromolar concentrations (HD-MB03 IC₅₀ = 0.13 µM), combined with demonstrated PAMPA-BBB permeability and a 44.3-fold therapeutic margin versus healthy neural stem cells [1]. For academic or biopharma teams evaluating HDAC6 as a therapeutic target in pediatric brain tumors, NN-390 provides a tool compound with disease-relevant validation that is absent for clinical candidates ricolinostat and citarinostat, which lack medulloblastoma efficacy data . The compound's plasma half-life (1.90 h in mice) and AUC exposure (2,523 h·ng/mL at 20 mg/kg IP) support once-daily dosing regimens in murine orthotopic xenograft models [2].

Structure-Based Drug Design

The availability of a high-resolution (1.6 Å) co-crystal structure (PDB 7UK2) makes NN-390 the rational starting point for structure-enabled HDAC6 inhibitor design [1]. Computational chemistry teams can directly use the NN-390-HDAC6 complex for docking, free energy perturbation, and molecular dynamics simulations to design next-generation analogs with improved residence time or subtype selectivity . Within the mercaptoacetamide chemotype, NN-390 is the only molecule with a public co-crystal structure, positioning it as the reference ligand for any SBDD campaign exploring tetrafluorobenzenesulfonyl cap group interactions with the HDAC6 L1–L2 loop pocket [2].

HDAC6 Selectivity Profiling

With experimentally confirmed >500-fold selectivity against HDAC1/2 and >100-fold against HDAC3/8/11, NN-390 serves as a high-confidence positive control for HDAC6-specific biochemical and cellular assays [1]. Its clean on-target pharmacology—α-tubulin acetylation at concentrations as low as 53 nM in medulloblastoma cells with no detectable histone H3 acetylation up to 211 nM—makes it superior to ricolinostat (~10–12-fold selective) and citarinostat (~18-fold selective) for experiments where Class I HDAC confounding must be rigorously excluded . Procurement of NN-390 for selectivity profiling panels enables accurate assignment of biological phenotypes to HDAC6 versus pan-HDAC inhibition [2].

CNS Pharmacokinetic Benchmarking

NN-390's comprehensively characterized pharmacokinetic profile—including matched comparative data against predecessor KT-531—provides a benchmark for PK/PD modeling of CNS-penetrant HDAC6 inhibitors [1]. The compound's improved PAMPA-BBB score and 2.8-fold plasma stability enhancement relative to earlier series members support its use as a reference compound in CNS exposure studies, particularly for programs targeting leptomeningeal metastases or other CNS malignancies where BBB penetration is a prerequisite for target engagement . Procurement of NN-390 alongside KT-531 enables head-to-head assessment of PK improvements attributable to the tetrafluorobenzenesulfonyl cap modification [2].

• [1] Nawar N, Bukhari S, Adile AA, et al. Discovery of HDAC6-Selective Inhibitor NN-390 with in Vitro Efficacy in Group 3 Medulloblastoma. J Med Chem. 2022;65(4):3193-3217. Primary publication containing all in vitro efficacy, PK, selectivity, cellular target engagement, and therapeutic margin data. View Source
• [2] 7UK2 crystal structure. PDB ID 7UK2. NN-390 bound to Danio rerio HDAC6 catalytic domain 2 at 1.6 Å resolution. Deposited 2022-03-31. View Source