Biological Activity and Structural Biology of Current KAT6A Inhibitor Chemotypes

Abstract

All lysine acetyltransferases (KATs) modulate biological outcomes through the acetylation of lysine side-chain amino groups facilitated by acetyl coenzyme A (AcCoA). KAT6A belongs to the class of MYST domain histone acetyltransferases (HATs), which had been regarded as undruggable. The first on-target KAT6A inhibitors with in vivo activity were reported in 2018, catalyzing intense industry interest in this enzyme as an oncology target. In this study, we experimentally evaluated representative KAT6A inhibitor chemotypes through resynthesis and comparative biochemical assays, cellular assays, and structural biology. We outline the recent history of each KAT6A inhibitor chemotype discovery, including SAR for potency, selectivity, and cellular activity. We extensively benchmark key compounds from each chemotype, augmented by new acylsulfonohydrazide analogues and a novel fused [1,2,4]­thiadiazine KAT6A inhibitor subclass, which we report here for the first time, along with co-crystal structures. Additionally, we report on the in vivo activity, pharmacokinetics, and toxicology profiles of these inhibitors.

Introduction

Histone acetylation states play an essential role in various cellular activities, including genome maintenance, DNA damage repair, cell cycle regulation, apoptosis, and broader biological functions. Many reports suggest that a disruption of the dynamic balance between acetylation/deacetylation in cells leads to various diseases, such as Parkinson’s disease, leukemia, and cancer. − Histone acetylation is governed by two enzyme families: histone acetyltransferases (HATs) and histone deacetylases (HDACs). All HATs are now defined as lysine acetyltransferases (KATs), as histone acetylation is synonymous with histone lysine acetylation. The KAT superfamily is responsible for all lysine acetylation, , and is known to play roles in various important cellular processes such as gene transcription, DNA damage recognition and repair, and DNA replication. KATs catalyze the transfer of an acetyl group from AcCoA to an ε-amino group of a histone lysine residue. − This modification removes the lysine residue’s positive charge and, in the case of histones, weakens interactions with negatively charged nucleosomal DNA and neighboring nucleosomes, resulting in a more open chromatin structure and facilitating transcription to take place. KATs that act on human histones and can be classified into six families (Table ), with MYST KATs, named after the initially identified members (MOZ, Ybf2, SAS2, and Tip60), , being the most prominent family among this group. ,−

1. KATs Types, Families, Subtypes, and Their Alternative Names ,,− .

Families	Subtypes	Alternative names
Type A:
• Located in the nucleus
• Play roles in the regulation of gene expression through acetylation of chromatin nucleosomal histones
• Some members contain a bromodomain
GNAT	KAT2A	GCN5
	Uniprot ID: Q92830
	KAT2B	PCAF
	Uniprot ID: Q92831
MYST	KAT5	Tip60 (60-kDa Tat-interactive protein), HTATIP, lysine acetyltransferase 5, protein 2-hydroxyisobutyryltransferase KAT5, protein crotonyltransferase KAT5, cPLA(2)-interacting protein.
	Uniprot ID: Q92993
	KAT6A	MOZ (monocytic leukemia zinc finger protein), MYST3, MYST-3, YBF2/SAS3, SAS2, TIP60 3 (60-kDa Tat-interactive protein 3), RUNXBP2 (runt-related transcription factor-binding protein 2), ZNF220 (zinc finger protein 220).
	Uniprot ID: Q92794
	KAT6B	MOZ2 (monocytic leukemia zinc finger protein 2), MORF (MOZ-related factor), MYST4, MYST-4, Qkf, YBF2/SAS3, SAS2, TIP60 4 (60-kDa Tat-interactive protein 4) and KIAA0383.
	Uniprot ID: Q8WYB5
	KAT7	HBO1 (histone acetyltransferase binding to ORC-1), MYST2, YBF2/SAS3, SAS2, TIP60 2 (60-kDa Tat-interactive protein 2), lysine acetyltransferase 7 and HBOa.
	Uniprot ID: O95251
	KAT8	Lysine acetyltransferase 8, MOF (males absent on the first), MYST1, YBF2/SAS3, SAS2 and TIP60 1 (60-kDa Tat-interactive protein 1) and (MYST-1; hMOF).
	Uniprot ID: Q9H7Z6
p300/CBP	KAT3B	EP300 and p300
	Uniprot ID: Q09472
	KAT3A	CBP, CREBBP
	Uniprot ID: Q92793
Transcription coactivators	KAT4	TAF1, TBP
	Uniprot ID: P21675
	KAT12	TIFIIIC90
	Uniprot ID: Q9UKN8
Steroid receptor coactivators	KAT13A	NCOA1, SRC1, BHLHE74
	Uniprot ID: Q15788
	KAT13B	NCOA3, SCR3, AIB1, ACTR
	Uniprot ID: Q9Y6Q9
	KAT13C	NCOA2, TIF2, bHLHe75, and GRIP1
	Uniprot ID: Q15596
	KAT13D	CLOCK
	Uniprot ID: O15516
Type B:
• Located in the cytoplasm
• Responsible for acetylating synthesized histones prior to their assembly into nucleosomes
• Does not contain bromodomain
Cytoplasmic	KAT1	HAT1
	Uniprot ID: O14929
	HAT4	NAA60
	Uniprot ID: Q9H7X0

The MYST family of KATs consists of five enzymes (KAT5, KAT6A, KAT6B, KAT7, and KAT8), with their defining features being the presence of the highly conserved MYST domain composed of an AcCoA binding motif and a zinc finger (Figure ). There continues to be further discoveries of new ways that the acetyltransferases impact cellular growth, differentiation, replication, and death, with some MYST KATs becoming prominent as disease targets, especially KAT5, KAT6A, KAT6B, and KAT7. − Some key roles that MYST KATs play in disease biology, particularly oncology, are summarized in Table . KAT6A, also commonly referred to as MOZ/MYST3, plays essential roles in normal hematopoietic stem cells and naturally suppresses cellular senescence. − The enzyme consists of 2004 amino acids with a double plant homeodomain (PHD), a MYST domain where the acetyl transfer takes place, an acidic region, a serine-rich domain, and a methionine-rich domain. −

1.

KAT6A domains and sites in comparison with other MYST KATs. ,,−

2. Summary of the Study of MYST KATs Biology and Association with Cancer to Date.

Enzymes	Residue length	Locations	Acetylation sites	Roles	Reported association with cancers
KAT5	513	Nucleus, cytoplasm	H2AZ lysine 719 and p53 at K120	Diverse roles in cellular processes, including the DNA damage response, the cell cycle, apoptosis, signaling, and transcriptional regulation. ,,	Inhibition of KAT5 inhibits cellular proliferation in a panel of prostate cancer cell lines and induces apoptosis through the activation of caspase 3 and caspase 9 in a concentration- and time-dependent manner. Loss of the enzyme causes cell cycle arrest, which is independent of p53, INK4A, and ARF.
KAT6A	2004	Nucleus	Acetylates itself, H2B, H3K14, H4 (K5, K8, K12 and K16) in vitro, and H3K9 and H3K23 in vivo	Regulation of transcription, chromatin organization, cell cycle progression, cell differentiation, signal transduction and response to cellular stress cues. Regulate multiple developmental processes, including hematopoiesis, neurogenesis, skeletogenesis, body segment identity, craniofacial development and heart development.	Highly expressed in several cancers, with the highest difference in gene expression in AML. Inhibition of KAT6A and KAT6B induces cell cycle exit and cellular senescence without causing DNA damage. KAT6A is amplified and/or overexpressed in 10–15% of breast cancer cases and represents a critical dependency in breast cancer cell lines with 8p11–12 amplification that also show elevated KAT6A expression. −
KAT6B	2073	Nucleus	H3K23
KAT7	611	Nucleus	H3K14	KAT7 is essential for the initiation of DNA replication by enabling prereplication complex (pre-RC) formation. Its phosphorylation by PKD1 at T97 and T331 increases stability by preventing ubiquitination, thereby promoting replication and cell proliferation. In addition, KAT7 plays a critical role in establishing central immune tolerance. A genome-wide CRISPR screen identified KAT7 as a driver of cellular senescence	Required for the maintenance of leukemia stem cells. Inhibition of KAT7 showed efficacy in a broad range of human cell lines and primary AML cells. Reducing KAT7’s HAT activity by crotonylation and acetylation diminishes the expression of genes involved in procentriole formation, thereby inhibiting colorectal cancer progression. KAT7 is also involved in the regulation of thymopoiesis. Its activity influences the development of T cells by modulating gene expression through histone acetylation, underscoring its importance in immune system maturation.
KAT8	458	Nucleus	H4K16	Embryonic development, critical for transcriptional activation of genes required for cell cycle progression.	Regulates androgen in prostate cancer cells, playing an essential role in the proliferation of cancer cells and colony formation.

Key historical developments in the investigation of the role of KAT6A in cancer are summarized in Figure . The first report of KAT6A’s involvement in the translocation t(8;16)­(p11;p13) of acute myeloid leukemia (AML) was published in 1996. In 2018, inhibition of KAT6A activity using small-molecule inhibitors was shown to induce cancer cell cycle exit and cellular senescence without causing DNA damage. This study reported the development of WM-1119, which validated KAT6A as an anticancer drug target and demonstrated that inhibition of the enzyme arrested progression in an Eμ-Myc transgenic mouse model of B-cell lymphoma. More recently, positive results from the Phase 1 clinical trial of KAT6A/B inhibitor PF-07248144 have provided clinical proof of concept targeting the enzyme as an oncology target. ,

2.

Reports related to the essential role of KAT6A and its involvement in cancer. ,− ,,−

The development of WM-1119 began with a high-throughput screen (HTS) of a library of some 243,000 structurally diverse “lead-like” compounds, which led to the identification of acylsulfonohydrazide CTX-0124143 (1), a reversible AcCoA-competitive inhibitor, as reported in 2011. , Medicinal chemistry optimization subsequently led to the development of the more potent inhibitors WM-8014 and WM-1119 with on-target activity both in vitro and in vivo. ,, Since then, intense interest has been growing in KAT6A as a drug target, with alternative inhibitor chemotypes emerging. These new chemotypes include fused [1,2,4]­thiadiazines, phenyl-″C5 heteroaryl″ sulfonamides, benzisoxazole sulfonamides, and acylsulfonamides with their representative compounds shown in Figure . −

3.

Main KAT6A inhibitor chemotypes and their representative molecules. ,−

Stimulated by the report of WM-1119 (3), further inhibitor discovery efforts led to the optimized lead CTx-648/PF-9363 (6), developed by Cancer Therapeutics Cooperative Research Centre (CTx) and Pfizer to furnish PF-07248144, which can be considered as the most clinically advanced KAT6A inhibitor. The compound entered Phase 1 clinical trials in 2020 for the treatment of advanced or metastatic solid tumors, including breast, prostate, and lung cancers. It was well tolerated and demonstrated strong KAT6A inhibition in peripheral blood mononuclear cells and tumors in the phase 1 dose escalation study. In the subsequent dose expansion study, PF-07248144 showed a tolerable safety profile and durable antitumor activity in heavily pretreated estrogen receptor-positive (ER+) HER2-negative (HER2−) metastatic breast cancer (mBC) with and without ESR1 or PIK3CA/AKT1/PTEN mutations. Building on these findings, PF-07248144 has progressed to a Phase 3 randomized trial, initiated in mid-2025, evaluating its efficacy and safety in combination with fulvestrant in patients with hormone receptor–positive (HR+)/HER2– advanced or metastatic breast cancer whose disease has progressed following prior CDK4/6 inhibitor therapy. More reports that suggest continued industry interest in KAT6A inhibitor development have recently been published, including MEN2312, developed by the Menarini Group, which entered a Phase 1 clinical trial for advanced breast cancer in 2024. Additionally, OP-3136, an orally available KAT6A/B inhibitor from Olema Oncology, entered Phase 1 trials in 2025 for advanced or metastatic solid tumors. ,

In this study, we experimentally evaluated representative KAT6A inhibitor chemotypes through synthesis, comparative biochemical and cellular assays, and structural biology. This included the resynthesis and testing of benchmark compounds under uniform assay conditions, accompanied by comparative structural biology. In addition, we report the synthesis and co-crystal structures of novel acylsulfonohydrazide analogues, as well as the discovery of a fused [1,2,4]­thiadiazine KAT6A inhibitor subclass. Together, these original experiments provide a consistent framework for evaluating leading KAT6A inhibitor chemotypes and their SARs.

Results and Discussion

Assays to Monitor KAT6A Inhibition

For acetyltransferase biochemical assays of bisubstrate enzymes like KAT6A, the catalytic mechanism, the specific construct of the enzyme used, the concentration of both substrates, and their respective Michaelis constants (K _M ) influence the measured half-maximal inhibitory concentration (IC₅₀) values. These factors make IC₅₀ values prone to variations across different studies and assay setups. Optimization of our current KAT6A biochemical assay confirmed that IC₅₀ values of the inhibitor vary depending on the concentrations of KAT6A, AcCoA, and H4 peptide used in the assay protocol (Supporting Information 1). The assays used to evaluate KAT6A modulators are listed in Table . During the development of the high-throughput screen (HTS) that led to the discovery of the first KAT6A inhibitor, acylsulfonohydrazide 1 (Figure ), a linear correlation was observed between the AlphaScreen signal intensity and KAT6A concentration within 2 and 20 nM range. , Higher enzyme concentrations were associated with an increase in the AcCoA-independent nonspecific signal, thereby reducing the signal-to-background (S/B) ratio. The K _M of KAT6A was reported to be 15 μM in the presence of 20 nM KAT6A, alongside the time dependency of the histone acetyltransferase (HAT) reaction. During optimization of the acylsulfonohydrazide chemotype that resulted in WM-8014 (2) and WM-1119 (3), , the AcCoA assay concentrations used were either 0.4 μM or 15 μM. , Testing the compounds at these two coenzyme concentrations provided evidence of competitive binding, as inhibitor potency decreased with increasing AcCoA levels. Several other KAT6A inhibitor chemotypes were subsequently reported by other groups, each evaluated using different assay techniques, AcCoA concentrations, KAT6A protein constructs, and histone peptides in their respective acetylation reactions. These assay conditions are summarized in Table . The table also includes the assay used to evaluate a group of KAT6A inhibitors reported to be irreversible, in which a spin column was employed to separate bound from unbound molecules prior to the enzymatic reaction.

3. Summary of Techniques for Assaying Small-Molecule KAT6A Modulator and Parameters Used in the Biochemical Acetylation Competition Assay against KAT6A Using AcCoA as the Acetyl Donor.

KAT6A modulator chemotypes	Assay	[AcCoA]	KAT6A construct and concentration	Histone peptide	ref
Acylsulfonohydrazides	AlphaScreen	15 μM	10 nM of MYST domain of KAT6A (residues 507–778)	50 nM N-terminal histone H4 peptide (residues SGRGKGGKGLGKGGAKRHRKV-GGK-biotin)	,
		0.4 μM and 15 μM			,
N-(2-oxoethyl)sulfanilamide-derivatives	Radioactive KAT6A inhibition assay	3 μM of radioisotope-labeled [3H] AcCoA	10 nM of MYST domains of KAT6A	2.5 μM of H3 peptide
[1,2,4]thiadiazines	AlphaScreen	1 μM and 10 μM	5 nM of MYST domain of KAT6A (residues 507–778)	100 nM of full-length recombinant H3.1
Phenyl-“C5 heteroaryl” sulfonamides		1 μM			− ,
Benzisoxazole sulfonamides
Fused isoxazolyl sulfonamide	TR-FRET	0.6 μM	MYST domain of KAT6A (residues 507–778)	200 nM of histone H4 peptide
Fused benzoisoxazolyl sulfonamides
Acylsulfonamides	TR-FRET	0.6 μM	KAT6A (residues 194–810)	220 nM of histone H4 peptide (residues	,
			0.6 nM of KAT6A (residues 194–810)	SGRGKGGKGLGKGGAKRHRKVLRDK-biotin)
Acylsulfonohydrazide and benzisoxazole sulfonamide KAT6A covalent inhibitors	AlphaLISA for inhibitory activity	0.4 μM	25 nM of KAT6A (residues 488–778)	200 nM of histone H3 peptide (residues 1–21)
	AlphaLISA after spin column to evaluate the reversibility/irreversibility		Prespin column : 100 nM of KAT6A (residues 488–778)	Post spin column : 200 nM of histone H3 peptide (residues 1–21)
			Post spin column : 100 nM of KAT6A (residues 488–778)

^aWhere available.

^bNot specified, but plausibly 0.6 nM.

Chemistry

Benchmark Compounds

In competition assays, the higher the concentration of the competing substrate, the less potent the test compound appears to be. Since AcCoA concentration is one of several assay parameters that vary across reports of KAT6A inhibitor potency, ,− , as shown in Table , the SAR needs to be interpreted in this context. To compare and contrast leading KAT6A inhibitor chemotypes reliably, it is preferable to use data generated under as uniform conditions as possible. For this purpose, in addition to developing our own acylsulfonylhydrazides and fused [1,2,4]­thiadiazines reported herein, it was necessary to resynthesize the key fused [1,2,4]­thiadiazine 4, benzisoxazole sulfonamides 6 and 7, and acylsulfonamide 8KAT6A inhibitors reported by CT_X, CT_X and Pfizer, and Bayer, respectively. ,, We also resynthesized the N-(2-oxoethyl) sulfanilamide-derived 9, as it is relevant to this study. Syntheses of benchmark compounds 4, 6, 7, 8, and 9 are provided in Supporting Information 1.

Acylsulfonyhdrazides 10a–e

Our acylsulfonohydrazide analogs 10a–e were synthesized by reacting appropriate acid hydrazides 11a–c with the appropriate sulfonyl chlorides or by the N,N,N′,N′-tetramethyl-O-(1H-benzotriazol-1-yl)­uronium hexafluorophosphate (HBTU) coupling method using carboxylic acid 12a–b (Scheme ). The hydrazide 11b and sulfonyl chloride 17, required for the synthesis of 10b, were prepared from 2-fluoro-5-hydroxybenzoic acid (13) and sodium 3-hydroxybenzenesulfonate (15), respectively (Scheme ).

1. Synthesis of Acylsulfonohydrazides 11a–e .

^a Reagents and conditions: (a) pyridine, R³SO₂Cl, 0 °C → rt., 6–82%; (b) i. MeCN, HBTU, rt. ii. N,N-ethyldiisopropylamine, rt. → 0 °C, 1 h iii. R⁴SO₂NHNH₂, reflux, 78–82%.

2. Preparation of Acid Hydrazide 9b (a) and Sulfonyl Chloride 17 (b) .

^a Reagents and conditions: (a) EtI, K₂CO₃, DMF, 50 °C, 35%; (b) NH₂NH₂.H₂O, EtOH, reflux, 67%; (c) K₂CO₃, DMF, 110 °C, 16 h; 1,3-dioxolan-2-one, 62%; (d) SOCl₂, DCM, 60 °C, 16 h

Fused [1,2,4]­thiadiazines 18a–d

2-Chlorothiophene-3-sulfonamides 20a and 20b were reacted with ethyl 2-ethoxy-2-iminoacetate, cesium carbonate and copper­(I) chloride in DMF at 80 °C to afford 21a and 21b, respectively These intermediates were subsequently heated with 2-(oxazol-2-yl)-2-phenylethan-1-amine in xylene to produce 18a and 18b (Scheme a). For synthesis of 18c and 18d, N-(tert-butyl)­thiophene-2-sulfonamides 23a and 23b were reacted with n-butyllithium and p-toluenesulfonyl azide, followed by hexadecyltributylphosphonium bromide and sodium borohydride, to yield the appropriate 3-amino-N-(tert-butyl)­thiophene-2-sulfonamides. These intermediates were converted to the hydrochloride salt 24a and 24b before reacting with ethyl carbonocyanidate to yield 25a and 25b, respectively. The final step involved a reaction in a microwave reactor with 2-(oxazol-2-yl)-2-phenylethan-1-amine to yield 18c and 18d (Scheme b).

3. Synthesis of Fused [1,2,4]­Thiadiazines 18a–d .

^a Reagents and conditions: (a) NH₃/EtOH, rt, 2h, 96%; (b) ethyl 2-ethoxy-2-iminoacetate, Cs₂CO₃, CuI, DMF, 80 °C, 25–30%; (c) 2-(oxazol-2-yl)-2-phenylethan-1-amine, xylene, 4 Å MS, 140 °C, 12–35%; (d) (CH₃)₃CNH₂, Et₃N, THF, 79%; (e) i. nBuLi, THF, −65 °C → −20 °C; ii. TsN₃, THF iii. TBHDPB, NaBH₄, THF iv. HCl, 50 °C, 30–61%; (f) ethyl carbonocyanidate, AcOH, 85 °C, 21–25%; (g) 2-(oxazol-2-yl)-2-phenylethan-1-amine, EtOH, MW 120 °C, 50%.

Structure–Activity Relationship and Selectivity

Here, we provide the key SAR of leading KAT6A inhibitor chemotypes, augmented by our acylsulfonohydrazides and fused [1,2,4]­thiadiazine analogues that complements the data package. Where data were available, the selectivity of our set of studied inhibitors toward different subtypes of the MYST KAT family is also summarized. In terms of KAT6A and KAT6B, although having independent in vivo functions, they share an identical MYST domain structure and highly conserved sequences across all functional domains. This allows inhibitors designed for KAT6A to also effectively target KAT6B, such compounds being referred to as KAT6, KAT6A/B or KAT6AB inhibitors. ,,−

Acylsulfonohydrazides

The SAR of acylsulfonohydrazide KAT6A inhibitors has been detailed by Leaver et al. and Priebbenow et al., summarized in Figure . Following the initial hit from HTS, CTx-0124143 (1), exploration on RHS and LHS (RHS: right-hand side; LHS: left-hand side of the molecule as drawn in Figure b) has led to the discovery of WM-8014 (2).

4.

Previous work on acylsulfonohydrazide KAT6A inhibitors SAR. One μM [AcCoA] was used in biochemical assay ,,− ,

WM-8014 (2) is relatively hydrophobic, rapidly metabolized by mouse and human liver microsomes, and has a high plasma protein binding. Attempts to optimize the PK properties by reducing its lipophilicity led to the discovery of the lead compound, WM-1119 (3). ,,, All members of this class of KAT6A inhibitors, including WM-8014 (2) and WM-1119 (3), were observed to be competitive with AcCoA, with the acylsulfonohydrazide moiety assumed to be anionic and recapitulating many of the active site interactions of the AcCoA pyrophosphate group. , Despite being a pyrophosphate mimetic, these compounds were sufficiently cell-permeable to dial in cell-based activity, a considerable feat given the poor track record for such behavior by pyrophosphate mimetics. ,

Regarding selectivity, WM-8014 (2) exhibited inhibition toward KAT6A with 4-fold, 28-fold and 43 more active against the enzyme than against KAT6B, KAT5 and KAT7, respectively. On the other hand, WM-1119 (3) was reported to be 1,100-fold and 250-fold more potent against KAT6A than against KAT5 or KAT7, respectively. Additionally, WM-1119 (3) did not exhibit affinity against a pharmacological panel of 159 diverse pharmacological targets.

In 2023, a series of N-(2-oxoethyl)­sulfanilamide-derived KAT6A inhibitors were reported based on acylsulfonohydrazide CTX-0124143 (1). The work mainly aimed to change the hydrazine N–N bond with an N–C bond, resulting in 9 (Figure ), which was reported to be more potent than 1 and exhibit good selectivity against other KATs. The reported biochemical IC₅₀ of 0.03 μM against KAT6A, was 25 times more potent against KAT7, 48 times more potent against KAT5, and surprisingly 277 times more potent against KAT6B, with no appreciable inhibition of KAT8.

5.

N-(2-Oxoethyl)­sulfonamide derived 9 KAT6A inhibitor developed based on 1. [AcCoA] = 1 μM.

A series of acylsulfonohydrazide KAT6A inhibitors has been reported to be irreversible, as represented by 26 (Figure ). For its irreversibility evaluation, a spin column was employed to separate bound from unbound molecules prior to the biochemical enzymatic reaction.

6.

Acylsulfonohydrazide KAT6A inhibitor 26 reported to be a covalent binder.

In our previous investigations on acylsulfonohydrazide-based KAT6A inhibitors, several key aspects of the arylsulfonyl moiety remained unexplored. , With a view to supplementing current knowledge on acylsulfonohydrazide SAR and structural biology, we designed, synthesized and tested new analogs 10a–e. These targets variously explore small alkyl ether extensions with varied hydrogen bonding capacities on the LHS while retaining either the previously reported 2-fluoro-3-methyl-5-(phenyl)-phenyl or 5-ethoxy-2-fluorobenzoic acid RHS substituents in 10a and 10b. Compound 10c-10e incorporate structurally divergent arylheterocycles in the LHS, in combination with either the previously reported 5-ethoxy-2-fluoro-3-methylphenyl and 3-fluoro-5-(pyridin-2-yl) RHS. , We also synthesized 9 to verify its literature-reported activity. The results of the biochemical assays for 9, 10a–e, and WM-1119 (3) are shown in Table . Of immediate note, 9 was essentially inactive, the only observable effect being an increase in the assay signal at the highest concentration test of 125 μM, rather than a decrease that would indicate KAT6A inhibition. Although the AcCoA concentration in the original report of 9 was 1 μM, compared to 10 μM in the current study (Table ), this difference would not be sufficient to account for such a loss of activity from the originally reported IC₅₀ value of 30 nM. Inactivity of 9 is unsurprising given that the sulfonamide would not be sufficiently acidic to mimic an anionic pyrophosphate group, and leaves doubt as to whether 9 is bioactive against KAT6A.

4. Inhibitory Potency of New Acylsulfonohydrazide 10a–e, Benchmark Compound WM-1119 (3) and N-(2-Oxoethyl)­sulfanilamide-Derived 9 in the AlphaScreen Biochemical Acetylation Competition Assay against KAT6A Using AcCoA as the Acetyl Donor.

^a10 μM [AcCoA]

^bNo inhibition occurred. An increase in the Alphascreen signal at the highest concentration test of 125 μM was observed.

In contrast, acylsulfonohydrazides 10a–10e all exhibited strong inhibitory activity against KAT6A, with 10c and 10e being the most active. The latter, derived from the replacement of 2-fluorophenyl group on the LHS of WM-1119 (3) with 3-fluoro-thiophen-2-yl, maintains potency with an IC₅₀ value of 0.088 μM, suggesting that this LHS part is available as an alternative moiety instead of the 2-fluorophenyl group. On the other hand, 10c embeds a 5-ethoxy-2-fluoro-3-methylbenzoyl moiety on the RHS and 1H-indole on the LHS, exhibiting strong KAT6A inhibitory activity with an IC₅₀ value of 0.045 μM, which is comparable to that of WM-1119 (3). For 10a, 10b, and 10d with the methoxymethyl, benzothiophene, and 2-hydroxyethoxy groups, respectively, adjacent to the sulfonyl moiety, potency was strong but slightly diminished, with IC₅₀ values of 0.11, 0.78, and 0.15 μM, respectively. Acylsulfonohydrazides 10c and 10d have the 5-ethoxy-2-fluoro-3-methylbenzoyl moiety on the RHS that we have shown previously to confer strong KAT6A inhibitory activity in analogous acylsulfonohydradizes bearing phenyl, 2-naphthyl, and 2-fluorophenyl groups, respectively, at the opposite ends. These prior analogues exhibited IC₅₀ values of 0.062 μM, 0.082 μM, and 0.074 μM, respectively. These were obtained using a lower AcCoA concentration of 0.4 μM, illustrating that in the case of 10c, the 1H-indole LHS confers better activity than these published compounds, even with a higher AcCoA concentration of 10 μM used in its current assay.

Fused [1,2,4]­thiadiazines

The discovery of acylsulfonohydrazide KAT6A inhibitors and validation of the enzyme as an anticancer drug target were followed by the development of fused [1,2,4]­thiadiazines and other inhibitor chemotypes. − The fused [1,2,4]­thiadiazine chemotype also showed selectivity for KAT6A, as reflected by the reported IC₅₀ values of 4 against KAT6A, KAT6B, KAT5, KAT7, and KAT8, which were 0.006, 0.059, 0.284, 0.237, and 24.5 μM, respectively. The SAR of the [1,2,4]­thiadiazine chemotype, as derived from the underlying patent, is shown in Figure . Both phenyl and pyridine rings were explored as fusions with the thiadiazine ring, although the bioactivities for the fused pyridine analogues were undisclosed.

7.

Previous works on fused [1,2,4]­thiadiazines KAT6A inhibitors SAR

Certain phenyl ring substituents on the X position of the fused system increased the activity. As detailed in Table , which summarizes diverse analogues with diverse substituents (4, 27b–s), some of which conferred favorable activity, with IC₅₀ ranging from 0.011 to 0.069 μM. The parent compound 27a with no 2-position substituent exhibited an IC₅₀ of 0.064 μM. Analogues 27b–j represent analogues with electron-withdrawing groups (EWGs) at the 2-position, with a simple acetyl group (27b) being the most active with an IC₅₀ of 0.014 μM. Interestingly, 27k with a hydroxyl group at the 2-position exhibited a similar level of activity, suggesting a potential role of hydrogen bond acceptor (HBA) interactions rather than EWG or electron-donating group (EDG) effects. Halogen substitution at the 2-position, except for fluorine, also improved activity, yielding single-digit nanomolar IC₅₀ values for compounds 4 and 27l–n. Various five-membered aromatic heterocycles were evaluated at the 2-position (27o–s), with IC₅₀ values ranging from 0.005 to 0.064 μM. The most potent analog in this set was the 4-(1H-1,2,3-triazol-4-yl) derivative, which exhibited an IC₅₀ of 0.005 μM.

5. Results of the AlphaScreen Biochemical Acetylation Competition Assay against KAT6A Using AcCoA as the Acetyl Donor of the Fused [1,2,4]­Thiadiazines Chemotype, Showing the LHS SAR of This Class with an Unsubstituted LHS Phenyl Analogue Comparator and the Representative Analogues with Different meta-Substituents .

Compound	R	IC50 (μM)
27a	H	0.064
27b	COCH3	0.014
27c	COOH	0.110
27d	COOCH3	0.120
27e	CONHCH3	0.030
27f	CON(CH3)2	0.047
27g	SO2CH3	0.043
27h	NHSO2CH3	0.054
27i	CF3	0.017
27j	CN	0.069
27k	OH	0.011
27l	F	0.074
		0.021 (eutomer)
		0.649 (distomer)
4	Cl	0.006
27m	Br	0.005
27n	I	0.005 a
		0.002 (eutomer)
		0.029 (distomer)
27o	thiazol-2-yl	0.024
27p	thiazol-4-yl	0.047
27q	1H-pyrazol-4-yl	0.041
27r	1-methyl-1H-pyrazol-4-yl	0.064
27s	1H-1,2,3-triazol-4-yl	0.005

^aAnalogues with the best activity are the ones with halogen except fluoro and 4-1H-1,2,3-triazole being the best meta substituent. The compounds were tested as racemic mixtures. The biochemical assay results of 27l and 27n enantiomers after chiral separation using supercritical fluid chromatography (SFC) purification methods are shown. 1 μM [AcCoA] was used in the biochemical assay.

^bSupercritical fluid chromatography (SFC) purification methods: Instrument – Waters SFC-80; Column: Chiralpak ADH (27l) and Lux C3 (27n) (250 × 20)­mm, 5 μm; Mobile phase: CO₂:MeOH (60:40); Total flow: 40 mL/min (27l), 60 mL/min (27n); Back pressure: 100 bar; Wavelength: 210 nm (27l), 304 nm (27n); Cycle time: 7 min (27l), 6 min (27n).

On the RHS, as summarized in Table , the amide NH was found to be essential for potency, with the N-methyl analogue 28 being approximately 100-fold less active, with a reported IC₅₀ of 0.43 μM, compared with its unsubstituted counterpart 27n with an IC₅₀ of 0.005 μM. Pairwise comparisons, as shown in Table , indicate that the oxazole ring could not be widely interchanged for other aryl groups, as replacement with a phenyl (29, IC₅₀ = 108 μM) or N-imidazole (30, IC₅₀ = 0.308 μM) led to activity loss of around 2-fold and 5-fold, respectively, compared with comparator 27a (IC₅₀ 0.064 μM). Substitution of the oxazole ring with imidaz-2-yl (31) led to a highly significant loss of activity (IC₅₀ = 99 μM). In contrast, replacement with a thiazole ring was well tolerated, as supported by observed activity of 32 (IC₅₀ 0.007 μM) compared with 4 and 38 with 27n (IC₅₀ 0.006 μM, 0.004 μM, and 0.005 μM, respectively). This contribution of the oxazole to inhibitory activity was further confirmed by comparing des-oxazole analogues 33 (IC₅₀ 0.203 μM) and 37 (IC₅₀ 0.098 μM) with their respective comparators 27m and 27n (both IC₅₀ = 0.005 μM).

6. Result of the Biochemical Acetylation Competition Assay against KAT6A Using AcCoA as the Acetyl Donor of Fused [1,2,4]­Thiadiazines Showing the Importance of Amide Hydrogen, Improvement of Activity by Substitution of the RHS Oxazole Ring, Effect of Substitution of RHS Phenyl Ring with Pyridine, and That Cyclohexyl Is a Well-Tolerated Replacement of Phenyl on RHS .

Compound	R1	R2	R3	R4	IC50 (μM)
27n	I	H	2-(oxazol-2-yl)	Ph	0.005
28	I	Me	2-(oxazol-2-yl)	Ph	0.427
27a	H	H	2-(oxazol-2-yl)	Ph	0.064
29	H	H	Ph	Ph	0.108
30	H	H	1H-pyrazol-1-yl	Ph	0.308
31	H	H	1H-imidazol-2-yl	Ph	99
4	Cl	H	2-(oxazol-2-yl)	Ph	0.006
32	Cl	H	thiazol-2-yl	Ph	0.007
27m	Br	H	2-(oxazol-2-yl)	Ph	0.005
33	Br	H	H	Ph	0.203
34	Br	H	pyridin-2-yl	Ph	0.011
35	Br	H	1H-pyrazol-1-yl	Ph	0.017
36	Br	H	Ph	Ph	0.682
37	I	H	H	Ph	0.098
38	I	H	thiazol-2-yl	Ph	0.004
39	I	H	1,3,4-thiadiazol-2-yl	Ph	0.008
40	I	H	5-oxo-4,5-dihydro-1,3,4-oxadiazol-2-yl	Ph	0.024
41	I	H	5-amino-1,3,4-oxadiazol-2-yl	Ph	0.075
42	I	H	5-(methoxymethyl)-1,3,4-oxadiazol-2-yl	Ph	0.024
43	I	H	3-(hydroxymethyl)-1,2,4-oxadiazol-5-yl	Ph	0.032
44	I	H	3-methyl-1,2,4-oxadiazol-5-yl	Ph	0.010
45	I	H	methoxyethyl	Ph	0.020
46	Cl	H	1H-pyrazol-1-y	Ph	0.017
47	Cl	H	1H-pyrazol-1-y	pyridin-2-yl	6.828
48	I	H	H	cyclohexyl	0.122

^aChiral compounds were tested as racemic mixtures. 1 μM [AcCoA] was used in the biochemical assay.

Analogues 39 and 44 were also very potent, with IC₅₀ values of 0.008 μM and 0.010 μM, respectively, testifying that certain other heterocycles were also favorable for activity, in this case, a thiadiazole and oxadiazole, respectively. The SAR was generally consistent across subseries, and pyrazole 35 (IC₅₀ 0.017 μM) was around 3-fold less active than oxazole 27m (IC₅₀ 0.005 μM), in line with the earlier comparison of compounds 30 and 27a. In this context, a pyrid-2-yl group (34, IC₅₀ = 0.011 μM) was also highly favorable and almost equivalent in activity to the oxazole comparator 27m. An exception to this trend was the RHS phenyl analogue 36, which, unlike the case of 29, was considerably disfavored in 36 (IC₅₀ = 0.682 μM). It is unclear why this is the case, but it is possible that the combination of a bulky and hydrophobic bromo group at the 2-position of the fused phenylthiadiazine ring and the bulky and hydrophobic RHS benzhydryl led to activity-limiting solubility or molecular bulk. Intriguingly, other relatively broad changes were well tolerated. These include the 5-oxo-4,5-dihydro-1,3,4-oxadiazol-2-yl (40), 5-amino-1,3,4-oxadiazol-2-yl (41), 5-(methoxymethyl)-1,3,4-oxadiazol-2-yl (42), 3-(hydroxymethyl)-1,2,4-oxadiazol-5-yl (43) and methoxyethyl (45) moieties, with IC₅₀ ranging from 0.020 to 0.075 μM. The replacement of the RHS phenyl ring with a 2-pyridyl group in the pairwise comparison of 46 and 47 led to a 400-fold activity loss. On the other hand, a pairwise comparison of a phenylethyl group with a cyclohexylethyl group on 38 and 48 revealed similar activity levels (IC₅₀ = 0.098 μM and 0.122 μM, respectively), suggesting cyclohexyl is a well-tolerated replacement of phenyl on RHS. This could be useful information because saturated rings are often preferred over aromatic rings in drug development. Following chiral separations of 27l and 27n, it was shown that their eutomers are 15–30 times more active than their corresponding distomers (Table ), which is later discussed further in the structural biology discussion.

To supplement current knowledge on SAR and structural biology around this chemotype, we synthesized and tested a series of compounds featuring an entirely new fused ring system, represented by 18a–d (Table ). These analogues incorporate a fused thiophene on the LHS instead of the fused LHS phenyl ring used in the original patent. We also made and tested benchmark 4 to enable side-by-side comparison, exhibiting 0.026 μM and 0.13 μM in our assay with 0.4 μM and 15 μM of AcCoA concentration, respectively. The new fused thiophenes showed remarkable regioselectivity, with 18b returning an IC₅₀ of 0.014 μM in 10 μM of AcCoA assay, while its regioisomeric counterpart 18d was more than 500 times less active, with an IC₅₀ of 6.1 μM despite being tested under more favorable conditions of lower coenzyme in the assay (0.4 μM AcCoA). The benchmark comparator 4 was less active than the highly potent 18b based on its IC₅₀ at a lower AcCoA level of 0.4 μM. For 18b, the loss of the chloro group led to a significant drop in potency, with des-chloro 18a returning an IC₅₀ of 0.27 μM in the lower AcCoA concentration assay. This again emphasizes the importance of a halogen atom on the LHS position for this chemotype, as previously shown in Table . Unexpectedly, the same modification had the opposite effect in the regioisomer series. The loss of the chloro group in 18d led to an increase in potency, with des-chloro 18c returning an IC₅₀ of 0.37 μM. This is attributed to the relatively long carbon–sulfur bonds that would cause the respective chloro groups to occupy distinctly different regions of space in the two regioisomeric thiophenes, with that in 18d likely encountering some degree of steric hindrance. The dramatic potency difference between regioisomers 18b and 18d underscores the sensitivity of this chemotype to subtle changes in substitution pattern, and suggests that precise vector orientation of substituents on the fused thiophene is critical for productive binding.

7. Inhibitory Potency of Our Fused [1,2,4]­Thiadiazines Analogs 18a–18d and Benchmark Compound 4 in the AlphaScreen Biochemical Acetylation Competition Assay against KAT6A Using AcCoA as the Acetyl Donor.

This result represents the first demonstration that a fused thiophene can effectively substitute for the fused phenyl ring in this scaffold, thereby expanding the structural diversity accessible within the fused [1,2,4]­thiadiazines chemotype. Notably, 18b was more potent than benchmark 4, even under higher AcCoA conditions.

Phenyl-“C5 Heteroaryl” Sulfonamides and Benzisoxazole Sulfonamides

A series of phenyl isoxazole sulfonamides was reported in 2020, represented by 5, which exhibited a KAT6A IC₅₀ value of 25 nM. Around the same time, a related series of benzisoxazole sulfonamides was reported, including CTx-648/PF-9363 (6), 7, and 49, each with a KAT6A IC₅₀ value of 0.006 μM. ,, Similar to acylsulfonohydrazides and fused [1,2,4]­thiadiazines, a hallmark of these compounds is the presence of an acidic sulfonamide NH group. A subsequent patent describes a novel anhydrous crystalline polymorph (Form 5) of 50, providing improved stability and manufacturability for pharmaceutical development.

Analogs containing a pyrazole at the benzylic position on the RHS and either dimethoxybenzene or cyclohexylamine on the LHS, represented in compounds 6 and 7, showed the best activity toward KAT6A. Benzisoxazole sulfonamide 6, which was developed from an initial HTS hit, was reported as a potent, selective and orally bioavailable inhibitor of KAT6A/B, with more than a 1,800-fold improvement in potency over the original HTS hit. The compound has reported K _i values for KAT6A, KAT6B, KAT5, KAT7, and KAT8 of 0.27 nM, 2.4 nM, 420 nM, 70 nM, and 670 nM, respectively. Inhibition of H3K23ac (KAT6A/B biomarker) by 6 was correlated with its antiproliferative effects and >11,000-fold improvement in potency against the biomarker compared to the original hit.

PF-07248144, likely belonging to this benzisoxazole sulfonamide chemotype, has completed its Phase 1 clinical trial in patients with ER+ HER2– metastatic breast cancer (mBC). ,,, Since the discovery of benzisoxazole sulfonamide KAT6A inhibitors, there have been at least eight new patents around this chemotype. ,,,, Representative compounds, including 51 from the fused isoxazolyl sulfonamide class, are shown in Figure . However, these compounds were reported to have less potent activities, as reflected by the IC₅₀ of compounds 51-53 in assays using lower concentrations of AcCoA. ,, Notably, compound 53 was developed by Aurigene Oncology and Olema Pharmaceuticals, linking this series to Olema’s clinical program, which has recently advanced a KAT6A inhibitor into clinical trials. In addition, In Silico Medicine disclosed related benzisoxazole sulfonamide inhibitors (52a and 52b) and subsequently licensed their patent portfolio to Menarini, thereby establishing a clear connection between these compounds and Menarini’s clinical program. , A group of benzisoxazole sulfonamide KAT6A inhibitors has also been reported to be irreversible, represented by 54. Furthermore, KAT6A degraders have recently been reported, with 55 being among the first class of KAT6A degraders in the patent.

8.

Phenyl-“C5 heteroaryl” and benzisoxazole sulfonamide and novel inhibitor classes around the chemotype. − ,,− ,,−

To obtain comparative biochemical assay, cell-based activity, and structural biology results, we resynthesized benchmark compounds 6 and 7. Compound 6 exhibited an IC₅₀ of 4.5 nM against KAT6A, while similarly, 7 has an IC₅₀ of 7.5 nM, as shown in Table . These results indicate that these lead compounds have better KAT6A inhibitory activity compared to the previously reported lead compounds from acylsulfonohydrazide and fused [1,2,4]­thiadiazine inhibitor chemotypes.

8. Inhibitory Potency of Benzisoxazole Sulfonamide Benchmark Compounds 6 and 7, and Acylsulfonamide Benchmark Compound 8 in the AlphaScreen Biochemical Acetylation Competition Assay against KAT6A Using AcCoA as the Acetyl Donor.

^a10 μM [AcCoA]

Acylsulfonamide

A series of subnanomolar KAT6A acylsulfonamide inhibitors has been published by researchers at Bayer (Figure ). ,, Like acylsulfonohydrazides, fused [1,2,4]­thiadiazines and benzisoxazole sulfonamides, the hallmark of this class is the presence of an acidic NH group. Ter Laak et al. provided more details behind the discovery of acylsulfonamide BAY-184 (59). The development began with high-throughput screening of approximately 4 million compounds using a KAT6A time-resolved fluorescence energy transfer (TR-FRET) biochemical assay. From the identified hit clusters, acylsulfonamide 60 was chosen based on their binding efficiency index (BEI), lipophilic efficiency, cellular activity, physicochemical properties and high solubility (Figure ). Acylsulfonamide 60 was also from the only cluster that was active in an estrogen receptor (ER) target gene reporter assay using MVLN cells. Hit optimization led to the discovery of the BAY-184 (59).

9.

Result of AlphaScreen biochemical acetylation competition assay against KAT6A using AcCoA as the acetyl donor of Bayer acylsulfonamides 8 and 56–59 as reported in their patent and publication documents. 0.6 μM [AcCoA] was used. It is assumed that 58 represents a racemic mixture. ,,

10.

Development of acylsulfonamide BAY-184 (59) from the HTS hit 60. 0.6 μM [AcCoA] was used in the AlphaScreen biochemical acetylation competition assay against KAT6A.

Regarding selectivity, acylsulfonamide 8 was 2.4-fold more potent against KAT6A than KAT6B. BAY-184 (59) returned IC₅₀ values of 71 nM, 83 nM, 14,300 nM, 1,070 nM toward KAT6A, KAT6B, KAT5 and KAT7, respectively, and was reported to be inactive against KAT8.

We resynthesized acylsulfonamide 8 to gain insight into the comparative biochemical and cell-based activity, as well as the structural biology of this inhibitor chemotype. We chose 8 based on its superior biochemical activity than 57, BAY-184 (59) and better activity in cell-based assays than 56 and 58. ,, The result of the biochemical assay for 8 is shown in Table . The observed IC₅₀ of 0.22 μM is notably higher than the previously reported value of 0.009 μM, likely due to the higher concentration of acetyl-CoA used in our competitive assay format. Based on the IC₅₀ values obtained from our comparative biochemical assay, it can be concluded that benzisoxazole sulfonamide 6 is the most potent KAT6A inhibitor among the benchmark compounds evaluated.

Physicochemical Properties

The physicochemical properties of the representative compounds from each KAT6A inhibitor chemotype are compared in Table . The molecular weight (MW) range of 389–448 is within the Lipinski rule criteria for all compounds. While the number of HBA becomes quite high for some compounds, such as 6 (10), the number of HBD remains low (1–2) across all compounds. For all compounds, the topological polar surface area (tPSA) values are sufficiently low, being below 140 Ų, to plausibly confer cell permeability, with a range of 88–111 Ų presented. , Most of the compounds are relatively acidic and likely to be deprotonated to varying extents at physiological pH, with calculated pK _a values of 3.5–6.8, conferring reduced clogD values for all compounds at physiological pH that are lower than the cLogP values in the range of 2.0–2.8, which are nonetheless also attractively low. The lipophilic ligand efficiency (LLE) values are all favorable, ranging from 5.9 to 7.8, where a value of at least 5.5 is generally considered highly druglike. Benzisoxazole sulfonamide 6 has the best LLE due to its high potency toward the KAT6A enzyme.

9. Physicochemical Properties of Key KAT6A Inhibitor Chemotypes .

Compound	MW (g/mol)	HBD	HBA	tPSA (Å2)	pKa (calc.)	cLogP	cLogD7.4	LLE
3	389	2	6	88	6.8	2.0	1.3	6.6
18b	437	2	8	109	4.0	2.4	0.4	7.5
6	444	1	10	111	3.5	2.6	0.6	7.8
8	448	1	8	94	3.7	2.8	0.8	5.9

^a10 μM [AcCoA] used in the AlphaScreen. biochemical acetylation competition assay against KAT6A using AcCoA as the acetyl donor. MW: molecular weight; HBD: hydrogen bond donor; HBA: hydrogen bond acceptor; tPSA: topological polar surface area; LLE: lipophilic ligand efficiency; HBD, HBA, pK _a, logP, logD values were calculated using the Percepta Platform of ACD/Laboratories; tPSA values were calculated using ChemDraw (version 21.0.0).

Structural Biology of KAT6A with Competitive Inhibitors

The co-crystal structure of MYST domain KAT6A in complex with AcCoA is shown in Figure . As will be discussed later, all reported KAT6A inhibitors to date bind within the upper region of the AcCoA binding site, mainly the pyrophosphate binding site, the AMP binding site and a hydrophobic and solvent exposed area behind it, which is not occupied by AcCoA on the LHS, and a portion of the pantothenic acid binding site on the RHS. The AcCoA pyrophosphate binding site on KAT6A, as shown in Figure , features four key HBDs contributed by the backbone of Arg655, Gly657, Gly659, and Arg660. These HBDs are well-positioned to interact with the pyrophosphate moiety, which contains four oxygen atoms acting as hydrogen bonding acceptors (HBAs). Previous X-ray crystallographic analysis of acylsulfonohydrazide KAT6A inhibitors confirmed that the inhibitors compete with AcCoA, with the acylsulfonohydrazide core acting as a pyrophosphate isostere within the binding pocket. ,,

11.

(a) Co-crystal structure of MYST domain of KAT6A with AcCoA (PDB ID: 2OZU, 2.3 Å) in ribbon representation of the protein; (b) surface representation; (c) Essential interaction of pyrophosphate moiety of AcCoA showing the role of 8 different hydrogen bonding features of the enzymes in this site, including a molecule of water; (d) structure of AcCoA.

As the pyrophosphate of AcCoA is bound as an anion, it stands to reason that the most potent KAT6A inhibitors also bind as an anion in the AcCoA binding site. Potentiometric titration data (Supporting Information 1) indicate that the acylsulfonohydrazides possess pK _a values in the range of 5.4–6.9, and together with the reported pK _a of 7.3 for WM-1119 (3), these findings suggest that they are at least partially deprotonated at physiological pH. The degree of deprotonation may vary among chemotypes due to the differences in pK _a range between them. An example is BAY-184 (59) with its reported pK _a of 4.55. The acidic character is likely critical for activity, and is likely the reason for the inactivity of 9 in our assay (Table ) since the replacement of hydrazide NH with CH most likely leads to a significant loss of sulfonamide NH acidity. We suggested that the reported potent bioactivity of 9 be viewed with great caution.

Structural analyses of KAT6A in complex with AcCoA and various inhibitors reveal the presence of a conserved water molecule as an additional feature in the active site, acting as a water bridge connecting the HBDs of the ligands with Tyr653, Lys656, Tyr658, and Gly659 of the enzyme, as shown in Figure in its interaction of KAT6A with its natural ligand. We confirmed that this water molecule is conserved in the complex of KAT6A with other inhibitor chemotypes, as shown in Figure . Furthermore, our observation of other MYST enzyme co-crystal structures in the database (RCSB PDB) revealed the same pattern of having conserved water in that position. This provides useful information for future receptor-based KAT inhibitor design. Structural water is known to be vital in mediating polar groups via hydrogen bonds and is essential for facilitating tight binding in some cases. It has also been suggested that water-mediated interactions can be as strong as direct interactions. , Considering that the conserved water in the pyrophosphate binding site is tightly located between the enzyme and ligand, mediating key interactions at the interface, it is considered essential to include the water molecule when attempting to model protein–ligand interactions of KAT6A or other KAT enzymes and in structure-based KAT6A inhibitor design.

12.

Overlay of co-crystal structures of MYST KATs with ligands showing the presence of conserved tight water in the pyrophosphate binding site. PDB ID: 2OZU, 2RC4, 6BA4, 8DD5, 6OIN, 6OIQ, 6OIR, 6OIO, 6OIP, 6OWH, 6OWI, 6PDD, 6PDE, 6PDF, 6PD9, 6PDA, 6PDB, 6PDC, 6PDG, 6PD8, 6BA2, 6CT2, 6MAK, 6MAJ, 7D0P, 7D0O, 7D0Q, 7D0R, 7D0S, 5GK9, 5WCI, 5J8C, 5J8F, 7CMR, 2GIV, 3TOA, 3TOB, 3QAH, 2PQ8 and 4DNC.

The interaction of KAT6A with acylsulfonohydrazide WM-1119 (3) and related analogues has been previously reported. ,, Here, we provide new insights and reveal additional RHS and LHS interactions within the KAT6A active site in the co-crystal structures of our divergent compounds 10a−e and which accounts for observed SAR (Figure ). Compounds 10a–e, like previously reported acylsulfonohydrazides WM-8014 (2) and WM-1119 (3), ,, form hydrogen bonds with Arg655, Gly657, and Arg660 of MYST^Cryst backbone atoms and the conserved water molecule inside the binding pocket, as well as additional hydrogen bonds with Ser690 located below the acylsulfonohydrazide core. The sulfonyl and carbonyl oxygens of the core engage in water-mediated hydrogen bonds with the backbone NH of Tyr653, Arg656, Tyr658, and Gly659. The benzylmethyl ether of 10a, 3-(2-hydroxyethoxy)­benzene (10b), the indolyl (10c), benzothiophenyl (10d), and the 3-fluorothiophenyl (10e) groups occupy the hydrophobic pocket on the LHS of the MYST^Cryst. Both the ether oxygen atom of 10a and the hydroxyl oxygen of 10b are at hydrogen bonding distances with the Arg660 side chain. The fluorine atom in the 3-fluorothiophenyl functional group in 10e forms a hydrogen bond with Ser690 in MYST^Cryst, while also forming an internal hydrogen bond with the hydrazide NH. This interaction was also found in the 2-fluorophenyl of WM-1119 (3), as reported in its co-crystal structure. The phenyl ring next to the core of 10a–e occupies the hydrophobic pocket on the RHS. Compounds 10b–d exhibit hydrogen bonding with the amino backbone of Ile 649 using their ethoxy oxygen atom, and 10e shows a similar interaction with the residue using its pyridine nitrogen, a feature also found in WM-1119 (3). The co-crystal structure of 10a–e suggests that a fluorine substituent at the 2-position can occupy the small hydrophobic pocket below the acylsulfonohydrazide core. In addition, the good potency reported for 10c suggested that the hydrophobic pocket below the phenyl ring can be addressed with fluorine or methyl at the 3-position. The ethoxy group can be an alternative to the WM-1119 (3) pyridine ring, which may be beneficial due to the lack of unnecessary positive charge of the pyridine nitrogen.

13.

New KAT6A-ligand complexes at high X-ray crystallographic resolution. Ribbon diagram showing: A) 10a bound to MYST^Cryst (PDB ID: 9OOC, 1.39 Å); B) 10b bound to MYST^Cryst (PDB ID: 9OOH, 2.05 Å); C) 10c bound to MYST^Cryst (PDB ID: 9OOD, 2.20 Å); D) 10d bound to MYST^Cryst (PDB ID: 9OOE, 2.10 Å); E) 10e bound to MYST^Cryst (PDB ID: 9OOF, 1.68 Å).

Here, we also provide insight into the comparative structural biology of key KAT6A inhibitor chemotypes. The co-crystal structures of acylsulfonohydrazide WM-1119 (3), fused [1,2,4]­thiadiazine 18c, benzisoxazole sulfonamides 6 and 7, and acylsulfonamide 8 in their interaction with MYST^Cryst are shown in Figure . Similar to acylsulfonohydrazide WM-1119 (3), other inhibitor chemotypes exhibit binding mechanisms that mimic pyrophosphate binding in their core by making hydrogen bonding networks with Arg655, Gly659, Arg660, and the conserved water molecule, as well as the water-mediated hydrogen bonds with the backbone NH of Tyr653, Lys656, Tyr658, and Gly659. As suggested in its design, pyrophosphate mimicry by acylsulfonohydrazide WM-1119 (3) is carried out by negatively charged acylsulfonohydrazide. All the negative charges of the inhibitors shown here are provided by their acidic sulfonamides. For acylsulfonohydrazide WM-1119 (3), HBAs of two phosphate group moieties of AcCoA are mimicked by the sulfonamide and acyl groups. Similarly, fused [1,2,4]­thiadiazine 18c provides the HBAs by the cyclic sulfonamide and the acyl group, while the negative charge is provided by deprotonation of the acidic NH in its ring. The co-crystal structure suggests a shift in the double bond, with the sulfonamide N–C bond (1.44 Å) being shorter than the lower N–C bond (1.54 Å), consistent with partial double bond character that could reflect redistribution of electron density to the sulfonamide oxygen atoms and the lower ring nitrogen atom to maximally engage the binding site. The nitrogen next to the sulfonyl also contributes as an additional HBA, forming hydrogen bonds with Arg655, Gly659, and the conserved water molecule in the pyrophosphate binding site. On the other hand, benzisoxazole sulfonamides 6 and 7 utilize the nitrogen atom of their oxazole ring as an HBA. The oxygen atom in this ring does not perform any polar interaction, suggesting that it is most likely to increase the acidity of the sulfonamide NH. Acylsulfonamide 8 also utilizes sulfonamide and the acyl group as its HBAs. Having only one atom between the two HBAs seems to be tolerated, as reflected by the activity of this chemotype and its co-crystal structure. This was expected as the pyrophosphate of AcCoA also only has an oxygen atom between its two phosphate groups. Interestingly, the co-crystal structure also managed to capture the interaction of the sulfonamide oxygen with Ser 690 at the bottom of the MYST^Cryst pyrophosphate binding site and the water bridge between the sulfonamide oxygen and the side chain guanidine of Ser660.

14.

Crystal structure of MYST^Cryst bound to A) AcCoA (PDB ID: 6BA4, 1.95 Å); B) acylsulfonohydrazide WM-1119 (3) (PDB ID: 6CT2, 2.13 Å); C) fused [1,2,4]­thiadiazines 18c (PDB ID: 9OOJ, 1.82 Å); D) benzisoxazole sulfonamide 6 (PDB ID: 9OO9, 2.20 Å); E) benzisoxazole sulfonamide 7 (PDB ID: 9OOA, 1.39 Å); and F) acylsulfonamide 8 (PDB ID: 9OOB, 1.83 Å).

These binding of acylsulfonohydrazide WM-1119 (3), fused [1,2,4]­thiadiazine (18c), benzisoxazole sulfonamides (6 and 7), and acylsulfonamide (8) cores demonstrated that KAT6A inhibition could also be achieved with isosteric varieties of pyrophosphate other than acylsulfonohydrazide, including fused bicyclic structures of [1,2,4]­thiadiazines, benzisoxazole sulfonamide and acylsulfonamide. Benzisoxazole sulfonamides 6 and 7, being the most active compounds in comparative biochemical assay (Table ), suggested the merit of the benzisoxazole sulfonamide as a pyrophosphate isostere.

In its co-crystal structure MYST^Cryst in complex with AcCoA, Ser690 has no contribution to polar interaction with the AcCoA pyrophosphate (Figure ). The co-crystal structure data indicates the evidence of the residue’s flexibility and its ability to form a hydrogen bond with the ring oxygen of the AcCoA ribose moiety. However, the importance of the MYST^Cryst Ser690 residue is observed in the interaction of the enzyme with the KAT6A inhibitor core. Figure shows a hydrogen bonding interaction between the amide hydrogen of fused [1,2,4]­thiadiazines 18c and the Ser690 oxygen. This also explains the importance of the amide hydrogen atom, as previously shown in a paired comparison of the activities of the free amide and N-methylated analogs in its SAR discussion (Table ). Ser690 also plays a role as an HBA in the polar interaction with acylsulfonohydrazide WM-1119 (3). The Ser residue contributes to key protein−ligand interactions with benzisoxazole sulfonamides 6 and 7 and acylsulfonamide 8. This residue and forms hydrogen bonds with the negatively charged sulfonamide nitrogen of benzisoxazole sulfonamide 6 and 7, and acylsulfonamide 8. The Ser residue also forms a hydrogen bond with the RHS methoxy oxygen, which is a feature shared by these three molecules. In the interaction of MYST^Cryst with acylsulfonamide 8, this serine residue forms an additional polar interaction with one of the sulfonamide oxygens. Another water bridge between the Ser with the side chain guanidine of Ser660 was also captured in its co-crystal structure.

On the LHS of the MYST^Cryst binding site, there is a hydrophobic pocket before an open and solvent-exposed area. This pocket was occupied by a fluorophenyl of acylsulfonohydrazide WM-1119 (3), a fused ring of fused [1,2,4]­thiadiazines 18c, a 2,6-dimethoxybenzene or a cyclohexylamine of benzisoxazole sulfonamide 6 and 7, and a 2-ethoxy-6-methoxyphenyl of acylsulfonamide 8 in their complexes with MYST^Cryst. The LHS fluorine atom of acylsulfonohydrazide WM-1119 (3) forms an internal hydrogen bond with the hydrazine hydrogen. Similarly, the negatively charged ring nitrogen of fused [1,2,4]­thiadiazines 18c forms an internal hydrogen bond with the amide of the molecule. This helps the amide keep the planar orientation relative to the core ring, favoring the interaction of the carbonyl oxygen with the HBD-rich region of the pyrophosphate binding site. The LHS methoxy group of benzisoxazole sulfonamide 6 and the ethoxy group of acylsulfonamide 8 provide additional polar interaction with Gly657.

On the RHS, all inhibitor chemotypes have an HBA to address the NH backbone of Ile649 in polar interaction, with a hydrophobic moiety to fill a pocket between this HBA feature and their core. An equivalent hydrogen bond acceptor is present in the natural ligand, served by the pantothenic acid moiety of AcCoA. In mimicking that feature, acylsulfonohydrazide WM-1119 (3) uses 3-fluoro-5-(pyridin-2-yl)­benzene, with its pyridine nitrogen serving as an HBA in a hydrogen bond with Ile649. Fused [1,2,4]­thiadiazines 18c has its 2-benzyloxazole with a methyl amine as a linker to its core, with its oxazole nitrogen taking the role of the HBA. Benzisoxazole sulfonamide 6 and 7 have their fused ring connected to a methyl-pyrazole to fill the hydrophobic pocket and utilize the pyrazole nitrogen as the HBA. Acylsulfonamide 8 has benzofuran and azetidine rings to fill the RHS hydrophobic pocket, with the benzofuran nitrogen forming a hydrogen bond with Ile649. The role of Ile649 as an HBD also explains the importance of the stereochemistry of fused [1,2,4]­thiadiazine compounds shown in Table , since the oxazole needs to be oriented into the RHS pocket to address the polar interaction, while the phenyl ring is positioned in the hydrophobic pocket next to the core. This would suggest that the eutomers of compounds 27l and 27n in Table are S-configured.

Structural Biology Considerations in Designing KAT Selective Inhibitors

Figure shows the residues that form the KAT6A binding pocket, comprising: the helix behind and the loop above the AcCoA pyrophosphate binding site, as well as the β-sheet on the RHS (pair 1, Ile647–Ile663); a helix at the front of the binding pocket (pair 2, Ser684–Ser697), loops further on the RHS (pair 3, Phe600–Leu601 and pair 4, Trp522–Tyr523); and a loop on the further LHS (pair 5, Lys659–Arg765). A comparison of the binding site residues of KAT6A with other MYST KATs is also shown in Figure . The residues responsible for forming polar interactions with KAT6A inhibitors are primarily from pair 1.

15.

A) Comparison of binding sites formed by residues of KAT6A/B with other MYST KATs. Residues contributing in hydrogen bond with KAT6A inhibitors and their equivalent residues in other KATs are colored in red; B) identified residue differences in the superimposed KAT5 (PDB ID: 2OUA, 1.85 Å), KAT6A (PDB ID: 2OZU, 2.30 Å), KAT7 (PDB ID: 6MAK, 2.13 Å) and KAT8 (PDB ID: 5WCI, 1.78 Å) structures. The colors for the carbon atoms in the bottom image are light blue for KAT5, green for KAT6A and AcCoA, gray for KAT7 and magenta for KAT8; C) KAT7 and H3K14ac (KAT7 PD biomarker) inhibitory activities and the co-crystal structure of KAT7 inhibitor WM-3835 (61, its carbon atoms are in green) bound to KAT7 (PDB ID: 6MAJ, 2.14 Å), with the activities and structure of KAT6A inhibitor WM-1119 (3, in black) are provided as a comparison. It is shown that WM-3835 (61) form polar interaction with Lys488 and Glu525 on the LHS binding site, which are unique to KAT7.

Notable differences were observed among KAT5, KAT6, KAT7, and KAT8 in residues where atoms orient toward the ligand, with no such differences between the highly similar KAT6A and KAT6B, which explains the limited selectivity of KAT6A inhibitors against KAT6B. ,,,− ,,,, These differences include Arg660 KAT6A/B, which is replaced by lysine in other MYST KATs; Ile649 in KAT6A/B, substituted by threonine in others, offering an additional HBD; and Gln654, which correspond to glutamine in KAT5 and KAT8, and methionine in KAT7. Met648 in KAT6A is replaced by leucine in other MYST KATs, and Ala693 is replaced by serine. Interestingly, KAT6A/B possesses a serine at position 697, while the equivalent residue is glutamine in KAT5, glutamic acid in KAT7, and tryptophan in KAT8.

An example of the strategic exploitation of these subtle differences is seen with WM-3835 (61), an acylsulfonohydrazide selective KAT7 inhibitor. The co-crystal structure of WM-3835 (61) bound to KAT7, along with its inhibitory activity against KAT7 and H3K14ac (KAT7 PD biomarker) is shown in Figure C. The structure reveals that WM-3835 (61) forms polar interactions with Lys488 and Glu525 on the LHS binding site of KAT7 via its 2-hydroxyl group. For comparison, the structure and activities of the KAT6A inhibitor WM-1119 (3), which lacks this LHS feature, are also provided.

Cell-Based Activity

As mentioned in the introduction, the first report of the involvement of KAT6A cancer was in the translocation t(8;16) (p11;p13) of acute myeloid leukemia in 1996. In addition, Yan et al. demonstrated that KAT6A depletion or inhibition by a small molecule inhibitor WM-1119 (3) suppresses growth, causes apoptosis and induces cell cycle arrest in human AML cell lines MOLM-13, MV4–11, and NOMO-1. The study also confirmed a dependency on KAT6A in an in vivo MOLM-13 xenograft model. Interestingly, MLLT1 (ENL) is among the top KAT6A codependencies in both the CRISPR and RNAi screen databases on the DepMap portal, and functions as a histone reader. − MLLT1 (ENL) binds to H3K9ac, H3K27ac, and other histone acyl modifications and is essential in maintaining oncogenic gene expression in AML. , A recent study also demonstrated that MOZ-TIF2, an oncogenic fusion protein of KAT6A (MOZ) and TIF2 found in AML, sustains high expression of developmental transcription factors implicated in leukemogenesis, binds promoter regions, and is enriched at active chromatin marked by H3K23 propionylation (H3K23pr). Pharmacological inhibition of KAT6A using WM-1119 (3), or targeted degradation of the MOZ-TIF2 protein disrupts AML maintenance by reducing leukemic self-renewal and promoting differentiation.

However, the initial target validation of KAT6A was first reported in 2018 using the murine Eμ-Myc lymphoma cell line EMRK1184 in vitro, which also demonstrated that KAT6A inhibition suppressed proliferation and induced cell cycle arrest through upregulation of Cdkn2a locus products. The selection of the B cell lymphoma cell line EMRK1184 was due to its expression of the Cdnk2a-locus-encoded ARF and wild-type p53.

More recently, the indication of focus for KAT6A inhibition is breast cancer, as reflected in ongoing clinical trials and recently published patents. ,, Amplification and overexpression of KAT6A in breast cancer have been reported to be associated with poor overall survival. , Sharma et al. compiled evidence of both KAT6A amplification in breast cancer cell lines and their dependency on the enzyme. This includes earlier findings showing that KAT6A is amplified and/or overexpressed in 10–15% of breast cancers, and that it represents a significant dependency in 8p11–12-amplified breast cancer cell lines with KAT6A overexpression. ,, Other evidence for KAT6A overexpression in breast cancer comes from the breast cancer patient data sets in TCGA and cBioPortal, which show that KAT6A gene amplification manifests in 8% of patients. Analysis of the Cancer Cell Line Encyclopedia (CCLE) also identified a subpopulation of breast cancer cell lines harboring KAT6A amplifications or overexpression. The tumors with KAT6A amplifications show significant correlations between KAT6A copy number and gene expression, and between KAT6A copy number, KAT6A gene expression, and protein levels. The dependency of breast cancer on KAT6A was based on the analysis of CRISPR knockout in the DepMap Portal, confirmed by a CRISPR/Cas9 competition assay for breast cancer cells and by antiproliferative assays using their benzisoxazole sulfonamide KAT6A inhibitor CTx-648/PF-9363 (6). ,

The cell-based activities of the KAT6A inhibitor main chemotypes to date are summarized in Table . As mentioned, the inhibition of KAT6A activity by acylsulfonohydrazide WM-1119 (3) induces cancer cell cycle exit and cellular senescence without causing DNA damage. Although 9 was reported to exhibit antitumor activity on four leukemia cell lines, its lack of KAT6A activity in our hands suggests that this cell-based activity may be nonspecific.

10. In Vitro Cell-Based Activities of the KAT6A Inhibitor Main Chemotypes.

Inhibitor chemotype	Compound	Cell-line	Description	ref
Acylsulfonohydrazide	WM-8014 (2) and WM-1119 (3)	E14.5 mouse embryonic fibroblasts (MEFs)	WM-8014 (2): IC50 = 2.4 μM; Induces senescence; no general cytotoxic effect.
			WM-1119 (3): Induce senescence cell cycle arrest at 1 μM with a phenotype similar to that seen upon treatment with WM-8014 (2).
		Murine Eμ-Myc lymphoma cell line EMRK1184	WM-8014 (2): IC50 = 2.3 μM
			WM-1119 (3) IC50 = 0.25 μM
	WM-1119 (3)	Human AML cell lines: MOLM-13, MV4;11 and NOMO-1 cells.	WM-1119 (3) significantly impaired colony formation in MOLM-13, MV4;11, and NOMO-1. Similar to genetic knockout, WM-1119 (3) treatment also increased CD11b expression, induced superoxide anion production (functional differentiation), and β-galactosidase activity (senescence) in MOLM-13 cells. Treatment of the compound almost completely abolished colony formation in RN2, ER-HOXA9, and ER-HOXB8 cells.
		RN2 (Mll-Af9; NrasG12D murine AML cell line), ER-HOXA9 (murine MLL-r-like AML model), and ER-HOXB8 (murine MLL-r-like AML model)
		MOZTIF2 LSKs	IC50 = 25 nM
Fused [1,2,4]thiadiazines	4	Histone H3 Lysine 23 acetylation biomarker assay	IC50 = 84 nM
		Cell line: U2OS (human osteosarcoma)
Benzisoxazole sulfonamide	CTx-648/PF-9363 (6)	Human breast cancer cells: ZR-75–1 (KAT6A Amp.), T47D (KAT6A Over-Exp), and MCF7 (KAT6A Low) antiproliferation assays.	The compound exhibited dose-dependent antiproliferative effects against ZR-75–1 and T47D, with minimal effects against MCF7.
		A panel of 60 breast cancer cell lines	A part of the breast cancer cell lines showed sensitivity to the compound with IC50 < 50 nM.
			The antitumor activity was enriched in the ER+ (luminal) breast cancer subtype.
	6	Histone H3 Lysine 23 acetylation biomarker assay	IC50 = 0.6 nM
		Cell line: U2OS (human osteosarcoma)
	55 (KAT6A degrader)	Western Blot KAT6A	KAT6A degradation >75% at 0.1 μΜ
		HCC1954 (ER–, KAT6A amp)	Treatment time: 24 h
		HiBiT system	DC50 < 0.01 μM; Dmax >75%.
		HeLa (HiBit-KAT6A)	Treatment time: 18 h
Acylsulfonamide	8, 56, 57 and 58	MVLN reporter assay	IC50 values:	,
		Proliferation assay: human ZR75–1, LCLC97TM1, and MDA-MB-436 (a human triple-negative breast cancer cell) proliferation assay.	8. MVLN: 39 nM; ZR75–1:72 nM; LCLC97TM1:213 nM; MDA-MB-436: >20 μM.
			56. MVLN: 74 nM; ZR75–1:89 nM; LCLC97TM1:277 nM; MDA-MB-436: >20 μM.
			57. MVLN: 542 nM; ZR75–1:644 nM; LCLC97TM1: > 20 μM; MDA-MB-436: >20 μM.
			58. MVLN: 648 nM; ZR75–1:3.3 μM; LCLC97TM1:2.8 μM; MDA-MB-436: >20 μM.
	BAY-184 (59)	MVLN reporter assay	IC50 = 168 nM
		Proliferation assay.	IC50/maximum inhibition:
		Human breast cancer cell lines/subtypes:	130 nM/58%
		ZR-75–1/ER-positive	210 nM/30%
		KPL-1/ER-positive	230 nM/53%
		EFM-19/ER-positive	450 nM/68%
		CAMA-1/ER-positive	460 nM/66%
		MCF7/ER-positive	550 nM/51%
		HCC1428/ER-positive	1200 nM/30%
		T47D/ER-positive	1250 nM/61%
		MFM223/triple negative	9990 nM/49%
		HCC1500/ER-positive	>20,000 nM/0%
		MDA-MB-436/triple negative	>20,000 nM/2%
		HCC38/triple negative	>20,000 nM/0%
		MIDA-MB-231/triple negative	>20,000 nM/22%
		MDA-MB-361/ER-positive	>20,000 nM/0%
		BT-549/triple negative	>20,000 nM/25%
		JIMT-I/HER2-positive	>20,000 nM/0%
		HCC1806/triple negative	IC50 = 796 nM; decreased the levels of H3K23ac.
		Human large cell lung carcinoma LCLC97TM1
		Histone H3 Lysine 23 acetylation biomarker assay	IC50 = 670 nM
		Cell line: ZR-75–1 (human breast cancer cells)

^aOnly dose–response curves (DRCs) were reported in the original study; IC₅₀ values were not provided.

The histone H3 lysine 23 acetylation (H3K23ac) biomarker assays performed in the human osteosarcoma cell line U2OS were reported for fused [1,2,4]­thiadiazines and benzisoxazole sulfonamide KAT6A inhibitors, with 4 and 6 exhibiting an IC₅₀ of 84 nM and 0.6 nM, respectively. , The biomarker assay was carried out for acylsulfonamide BAY-184 (59) in human breast cancer cells ZR-75–1, with an IC₅₀ of 670 nM.

In the case of breast cancer, estrogen receptor-positive (ER+) subtypes have been shown to be more sensitive to KAT6A inhibition. KAT6A directly regulates the transcription of ESR1, the gene that encodes ERα. Furthermore, KAT6A knockdown using shRNA led to decreased ERα levels in ER+ breast cancer cells, highlighting KAT6A’s role in controlling ERα-driven proliferation in these cells and accounting for the sensitivity of ER+ breast cancer cells toward KAT6A inhibitors. ,,,,

Benzisoxazole sulfonamide CTx-648/PF-9363 (6) demonstrated potent antiproliferative activity in KAT6A-high, ER+ human breast cancer cell lines, including ZR-75–1 and T47D, while showing minimal activity in the KAT6A-low ER+ MCF7 cell line. The antiproliferative effect of 6 was correlated with the inhibition of H3K23ac (KAT6A/KAT6B PD biomarker).

Acylsulfonamide BAY-184 (59) inhibited the proliferation of human ZR-75–1 breast cancer and showed activity against several other breast cancer cell lines (Table ), most of which were ER+. Repression of ESR1 transcription, which is controlled by KAT6A in breast cancer cells, by 59 was quantified using an MVLN reporter assay. Interestingly, 59 also exhibited activity against LCLC97TM1 human lung cancer cell lines, which seems to be independent of the estrogen receptor. The activity was associated with a decrease in the levels of H3K23ac. Acylsulfonamide 8 (Figure ) was the most active compound from this chemotype, which was active in the MVLN reporter assay and proliferation assay using ZR75–1 (a human breast carcinoma) and LCLC97TM1 (a human large cell lung carcinoma) (Table ).

Degradation of KAT6A by treatment of benzisoxazole sulfonamide KAT6A bifunctional degrader 55 was confirmed by Western Blot KAT6A degradation assessment using HCC1954, a human ER- breast cancer cell line characterized by KAT6A gene amplification, which showed that the PROTAC molecule induced >75% KAT6A degradation at 0.1 μΜ concentration. The degradation activity was evaluated by measuring levels of KAT6A protein using the HiBiT system in HeLa cells with a HiBit-tagged KAT6A, which showed that 55 had a DC₅₀ of less than 0.01 μM and a maximum degradation (D _max) of more than 75%. In addition, a study by Ao et al. revealed that a CDK9 inhibitor-based PROTAC was able to lessen the lysine acetyltransferase KAT6A in a proteomic analysis. Their streptavidin immunoprecipitation (IP) assay showed a direct interaction between KAT6A and iCDK9, which suggested that KAT6A is a potential nonkinase target of iCDK9.

In our study herein described, we carried out a benchmark cell-based assay to evaluate the activity of key KAT6A inhibitors, including WM-1119 (3), fused thiadiazine 4 and the more potent analogue 18b, benzisoxazole sulfonamides CTx-648/PF-9363 (6) and 7, and acylsulfonamide 8, using the human ZR-75–1 and T-47D cancer cell lines. Sharma et al. previously reported the dependence of the growth of these lines on KAT6A, along with the potent antiproliferative activity of benzisoxazole sulfonamide 6 using the same model. This benchmark assay provided insight into the comparative cell-based activities of current KAT6A inhibitor classes and revealed the previously unreported antiproliferative activity of fused [1,2,4]­thiadiazines KAT6A inhibitors.

The results of the cell proliferation assay of acylsulfonohydrazide WM-1119 (3), fused [1,2,4]­thiadiazines 4 and 18b, benzisoxazole sulfonamides 6 and 7, and acylsulfonamide 8 against human ZR-75–1 and T-47D breast cancer cell lines are shown in Figure . A similar activity pattern was observed in both cell lines, with benzisoxazole sulfonamides 6 and 7 displaying the highest potency. Compound 6 exhibited IC₅₀ values of 0.94 nM and 2.6 nM against ZR-75–1 and T-47D, respectively, while 7 showed IC₅₀ values of 50 nM (ZR-75–1) and 12 nM (T-47D).

16.

Effects of the indicated compounds on proliferation of A) ZR-75–1 and B) T-47D human breast cancer cells, along with their corresponding IC₅₀ values.

Acylsulfonohydrazide WM-1119 (3) demonstrated comparable activity across both cell lines, with IC₅₀ values of 64 nM for ZR-75–1 and 66 nM for T-47D cells. Fused [1,2,4]­thiadiazines 4 and 18b were the least active, showing IC₅₀ values of 1,600 nM and 1,200 nM against ZR-75–1, and 1,900 nM and 1,500 nM against T-47D, respectively. Acylsulfonamide 8 exhibited moderate activity, with IC₅₀ values of 150 nM and 160 nM against ZR-75–1 and T-47D, respectively.

In Vivo Activity, Pharmacokinetics, and Toxicology

Acylsulfonohydrazides WM-8014 (2) and WM-1119 (3) have been shown to arrest lymphoma progression in mice and induce cell cycle exit and oncogene-induced cellular senescence in a zebrafish model of hepatocellular carcinoma without damaging DNA. The absence of DNA damage is an important attribute since anticancer chemotherapeutics have a tendency to damage DNA and are limited in their utility by so doing. −

Permeability and bioavailability seemed not to be key liabilities for acylsulfonohydrazide KAT6A inhibitors. WM-8014 (2) was reported to have high permeability with A-B P_app of 77.5 × 10^–6 cm/s in Caco-2, while WM-1119 (3) showed P_app of 60 × 10^–6 cm/s. , That said, it has been suggested based on bidirectional permeability studies reported by workers at Bayer, that efflux is a potential issue with this chemotype. WM-1119 (3) demonstrated oral bioavailability of 56% in rats, while 9 showed similar oral bioavailability (55%) following oral administration in rats. Both WM-8014 (2) and WM-1119 (3) exhibit high plasma protein binding (PPB), with binding values greater than 99.9% and up to 98.5%, respectively. In addition, WM-8014 (2) was rapidly metabolized in the presence of NADPH, with degradation half-lives of ≤ 32 min in both mouse and human liver microsomes. Significant non-NADPH-dependent metabolism of the compound was also observed, particularly in mouse liver microsomes. Medicinal chemistry optimization addressed this problem of rapid metabolism while maintaining activity with the discovery of acylsulfonohydrazide 62 and 63, which displayed improved metabolic stability with degradation half-lives of 61 and 163 min in human liver microsomes, respectively (Figure ).

17.

Acylsulfonohydrazides KAT6A inhibitor with improved microsomal stability.

The in vivo antitumor efficacy of benzisoxazole sulfonamide CTx-648/PF-9363 (6) was evaluated in ER+ human breast cancer xenograft model using ZR-75–1 cells subcutaneously implanted in immunodeficient NSG mice. The compound exhibited dose-dependent antitumor activity, inducing complete tumor stasis at a dose of 0.03 mg/kg. Tumor regressions resulting in complete response (CR) were observed at 0.3 and 5 mg/kg doses of 6, administered orally once daily (PO, QD). Investigation of the PD biomarker response of 6 in vivo was evaluated in NSG mice subcutaneously engrafted with ZR-75–1 tumors, dosed with 0.1 or 3 mg/kg once a day (QD) for 10 days. The study showed that the systemic free levels of 6 were more than the concentrations required to deplete the H3K23ac biomarker in vitro in cells (IC₅₀ = 0.85 nM). The analysis of PD biomarker H3K23Ac modulation in tumors highlighted the improvement of in vivo pharmacokinetic properties of 6 over the previous generation of acylsulfonohydrazide KAT6A inhibitor chemotypes.

Pharmacokinetic evaluation showed that 6 had intrinsic clearance values in human, rat, and mouse hepatocytes of 0.37, <18 and <33 mL/min/kg, respectively, and demonstrated 92% of oral bioavailability in rats. Benzisoxazole sulfonamide 52a and 52b, despite its lower activity toward KAT6A compared to CTx-648/PF-9363 (6), was reported to have a better permeability in the Caco-2 assay. Compound 52a exhibited a mean A-B P_app of 8.59 × 10^–6 cm/s and an efflux ratio of 5.06, while compound 52b displayed a mean A→B Papp of 3.38 × 10^–6 cm/s and an efflux ratio of 6.28. In contrast, CTx-648/PF-9363 (6) showed markedly poorer permeability, with an A-B P_app of 0.63 × 10^–6 cm/s and an efflux ratio of 42.4 measured for CTx-648/PF-9363 (6). , In the dose escalation data from the Phase 1 clinical trial of PF-07248144, the noted dose-limiting toxicity (DLT) was neutropenia. The reported treatment-related AEs (TRAEs) were dysgeusia, anemia, neutropenia, thrombocytopenia, diarrhea, leukopenia, fatigue, and elevated aspartate aminotransferase.

BAY-184 (59) inhibited the growth of ZR-75–1 xenografts in NMRI nude mice significantly at all doses, with complete growth inhibition at a dose of 150 mg/kg administered via oral gavage twice daily (2 QD). A significant reduction in H3K23 acetylation levels within tumor tissues at 1 and 3 h post administration in the 100 mg/kg 2 qd treatment group. The sodium salt of BAY-184 (59) significantly inhibited the growth of LCLC-TM971 xenografts in NMRI nude mice, achieving complete growth inhibition at a dose of 150 mg/kg, once daily (QD). In human liver microsome, BAY-184 (59) showed low blood clearance, CL_blood and F _max values of 0.77 L/h/kg and 41%, respectively. In human hepatocytes, the corresponding values were 0.99 L/h/kg and 25%, respectively. Permeability through the Caco-2 cell was high, with P_app A-B value of 1.46 × 10^–5 cm/s, P_app B-A of 1.00 × 10^–5 cm/s, and an efflux ratio of 0.69, indicating minimal efflux. The compound showed CL_blood value of 0.34 L/h/kg in rats and 0.56 L/h/kg in mice, with moderate oral bioavailability of 60% in rats.

BAY-184 (59) inhibited CYP2C9 with an IC₅₀ of 0.48 μmol/L and none observed (IC₅₀ > 20 μmol/L) for CYP1CA2, CYP2C8, CYP2D6, and CYP3A4. Acylsulfonamide BAY-184 (59) (Figure ) is associated with mutagenic liability, as it was positive in the Ames test. This mutagenicity was suggested to arise due to the formation of the reported mono-N-methyl and aniline metabolites via N-demethylation. Mutagenicity data was not available for 8 in its patent document. However, other acylsulfonamides, including 57 and several analogues, were reported to be nonmutagenic, in contrast to 59. , This supports the hypothesis that the mutagenicity of 59 might arise from its dimethylaniline moiety.

18.

Mutagenic activities by Ames test of acylsulfonamide KAT6A inhibitors. ,,

Conclusions

The recent validation of KAT6A as an anticancer drug target has catalyzed intense industry interest in the discovery of inhibitors targeting this enzyme. Here we contribute to this growing field by providing a review of the SARs, a comparative biochemical activity, structural biology, and cell-based antiproliferative activity, along with additional insights into the in vivo efficacy, pharmacokinetics, and toxicological profiles of key KAT6A inhibitors. We also introduce our new acylsulfonohydrazide analogues and a novel subclass of fused [1,2,4]­thiadiazines, including analogue 18c, which demonstrated superior biochemical activity compared to the benchmark compound 4 from its class.

Our structural biology analysis reveals that all reported KAT6A inhibitors act as pyrophosphate isosteres, consistently employing an acidic sulfonamide motif to mimic the anionic phosphate group of AcCoA. These inhibitors engage four key hydrogen bond donors within the KAT6A active siteArg655, Gly657, Gly659, and Arg660and a conserved water molecule that bridges Tyr653, Lys656, Tyr658, and Gly659, further stabilizing the interaction. The study highlights the use of chemically diverse substituents at the RHS, occupying the AcCoA pantothenic acid binding site, and the LHS, extending into a hydrophobic and solvent-exposed area adjacent to the core.

Drug discovery efforts in this field have achieved significant progress, yielding a series of KAT6A inhibitor chemotypes that offer valuable insight into novel pyrophosphate isosteres, supported by biochemical, cell-based, in vivo, and clinical studies. Modalities other than inhibition are being reported, and although reported data to date are insufficient to support the merit of these approaches, mechanisms involving covalent binding and degradation are being investigated. ,

Ultimately, based on the strength of the preclinical data, a selective KAT6A/B inhibitor PF-07248144 entered Phase 1 clinical trials in 2020 for the treatment of advanced or metastatic solid tumors, including breast, prostate, and lung cancers. Industry interest in KAT6A inhibitor development continues to grow, with MEN2312 entering Phase 1 clinical trials for advanced breast cancer in 2024, and OP-3136 entering Phase 1 trials for advanced or metastatic solid tumors in 2025. − PF-07248144 has since completed Phase 1 trials in patients with advanced or metastatic ER+ HER2– breast cancer, demonstrating promising clinical outcomes, , and has now advanced to a Phase 3 trial, thereby establishing clinical proof of concept for targeting KAT6A as an epigenetic therapy for ER+ HER2– metastatic breast cancer (mBC).

Experimental Section

General Information

Solvents were dried and distilled prior to use by literature methods or purchased in their anhydrous form from Merck. All chemical reagents were purchased from commercial sources and used as received unless otherwise indicated. Thin layer chromatography was carried out on analytical Merck TLC silica gel 60 F254, viewed by UV light (254 nm). Column chromatographic purification was performed on SiliCycle SiliaFlash P60 40–63 μm (230–400 mesh) 60 Å Irregular Silica Gels.

LC-MS

LCMS was performed on an Agilent 6120 Single Quadrupole LCMS coupled with an Agilent 1260 Infinity UHPLC, 1260 Infinity G1312B binary pump, 1260 Infinity G1367E HiP ALS autosampler, and a 1290 Infinity G4212A photodiode array detector. Analyses were performed using a Poroshell 120 EC-C18 (3.0 × 50 mm, 2.7 μm) column at a temperature of 35 °C, with an injection volume of 2 μL. Gradients utilizing 0.1% formic acid in water (solvent A) and 0.1% formic acid in MeCN (solvent B) were used, typically 5–100% of solvent B in solvent A over 2.5 min, then maintained at 100% for 3.8 min. UV detection was performed at 254 and 214 nm, while MS detection was done using either APCI or ESI, with a drying gas temperature of 350 °C, and a capillary voltage of 3000 V (positive and negative).

HPLC

All compounds are >95% pure by HPLC analysis. Compound purity was determined on an Agilent 1260 Infinity analytical HPLC equipped with a G1322A degasser, G1312B binary pump, G1367E HiP ALS autosampler, G1316A thermostatted column compartment, and G4212B DAD detector. Analyses were performed using an Agilent Zorbax Eclipse Plus C18 Rapid Resolution (4.6 × 100 mm, 3.5 μm) column at a temperature of 35 °C, with an injection volume of 2 μL. Gradients utilizing 0.1% TFA in water (solvent A) and 0.1% TFA in MeCN (solvent B) were used, typically 5–100% of solvent B in solvent A over 9 min, then maintained at 100% for 1 min. UV detection was performed at 254 and 214 nm.

NMR

NMR spectra were obtained at 400.13 MHz (¹H) and 100.6 MHz (¹³C) on a Bruker Ultrashield 400 spectrometer. The deuterated NMR solvents were obtained from Cambridge Isotope Laboratories. All chemical shifts were measured in parts per million (ppm) with the following solvent peaks as an internal standard of deuterated chloroform (CDCl₃) (7.26 ppm for ¹H and 77.16 ppm for ¹³C), deuterated methanol (CD₃OD) (3.31 ppm for ¹H and 49.0 ppm for ¹³C), deuterated dimethyl sulfoxide (DMSO-d ₆) (2.50 ppm for ¹H and 39.52 ppm for ¹³C) and deuterium oxide (D₂O) (4.79 ppm for ¹H). The signals were reported as the following convention: chemical shift (δ), multiplicity, and coupling constant (J Hz).

HRMS

HRMS was performed on an Agilent 6224 TOF LC/MS equipped with an Agilent 1290 Infinity HPLC. All data were acquired, and reference mass was corrected via a dual-spray electrospray ionization (ESI) source. Each scan or data point on the total ion chromatogram (TIC) was an average of 13,700 transients, producing a spectrum every second. Mass spectra were created by averaging the scans across each peak and background subtracted against the first 10 s of the TIC. Acquisition was performed using the Agilent Mass Hunter Data Acquisition software version B.05.00 Build 5.0.5042.2 and analysis was performed using Mass Hunter Qualitative Analysis Version B.05.00 Build 5.0.519.13. MS detection was done using ESI with drying gas flow of 11 L/min, capillary voltage of 4000 V, a 160 V fragmentor, skimmer of 65 V, octupole RFV of 750 V, scan range of 100–1500 m/z, internal reference ions in positive ion mode at m/z of 121.050873 and 922.009798. Prior to HRMS analysis, samples were separated on an Agilent Zorbax SB-C18 Rapid Resolution HT (2.1 × 50 × 1.8 mm) column. Gradients utilizing 0.1% formic acid in water (solvent A) and 0.1% formic acid in MeCN (solvent B) were used, with 5–100% of solvent B in solvent A over 3.6 min.

Compound Synthesis

General Procedure I for Hydrazide and Sulfonyl Chloride Coupling (10a–c)

To a hydrazide solution (1.0 equiv) in pyridine (0.3 M) at 0 °C was added the appropriate sulfonyl chloride (1.2 equiv). The resulting reaction mixture was allowed to stir for 2 h at room temperature after which the reaction mixture was poured onto water. The desired organics were extracted with EtOAc and the combined organic layers were washed with 1 M HCl, water and then dried over MgSO₄. The organics were then filtered and concentrated in vacuo to yield crude was purified by flash chromatography using silica gel and 30% EtOAc/cyclohexane to yield the compound of interest.

General Procedure II for Acid and Sulfonyl Hydrazide Coupling (10d–e)

To a stirred solution of the carboxylic acid in acetonitrile (0.1 M) was added HBTU (1.0 equiv) and the resulting reaction mixture was cooled to 0 °C. N,N-diisopropylethylamine (1.0 equiv) was added and the reaction mixture was allowed to stir at 0 °C for 1 h. The sulfohydrazide (1.2 equiv) was then added and the reaction mixture was heated to reflux overnight. The reaction mixture was then cooled to room temperature and was dry loaded onto silica in vacuo before being purified by column chromatography using silica gel as the stationary phase and a mixture of petroleum benzine/EtOAc as the eluent.

N′-(4-Fluoro-5-methyl-[1,1’-biphenyl]-3-carbonyl)-3-(methoxymethyl)­benzene-sulfonohydrazide (10a)

White amorphous solid (82%); ¹H NMR (400 MHz, DMSO-d ₆) δ 10.66 (d, J = 3.5 Hz, 1H), 10.18 (d, J = 3.5 Hz, 1H), 7.85 (s, 1H), 7.81 (d, J = 6.5 Hz, 1H), 7.73 (d, J = 4.8 Hz, 1H), 7.64 (d, J = 7.2 Hz, 2H), 7.56 (d, J = 6.6 Hz, 2H), 7.49 (t, J = 7.6 Hz, 2H), 7.42–7.37 (m, 2H), 4.46 (s, 2H), 3.23 (s, 3H), 2.31 (s, 3H); ¹³C NMR (100 MHz, DMSO-d ₆) δ 163.4, 139.8, 139.4, 139.0, 136.4, 132.8, 132.3, 129.5, 129.3, 128.2, 127.2, 127.1, 126.8, 126.3, 125.6, 73.1, 58.1, 14.7; MP:193.1–195.6 °C; HPLC: t _R = 6.3 min, > 99% purity (254 nm).

N′-(5-Ethoxy-2-fluorobenzoyl)-3-(2-hydroxyethoxy)­benzenesulfonohydrazide (10b)

White amorphous solid (58%); ¹H NMR (400 MHz, DMSO) δ 10.54 (bs, 1H), 10.13 (bs, 1H), 7.47 (t, J = 6.7 Hz, 2H), 7.40 (s, 1H), 7.30–7.09 (m, 2H), 7.10–6.95 (m, 1H), 6.84 (dd, J = 5.3, 3.2 Hz, 1H), 4.48–4.15 (m, 4H), 3.99 (q, J = 6.9 Hz, 2H), 1.30 (t, J = 6.9 Hz, 3H); ¹³C NMR (100 MHz, DMSO) δ 163.2, 158.4, 153.7, 154.8, 140.6, 130.6, 122.6, 120.8, 120.0, 119.2, 117.6, 114.9, 113.9, 67.2, 64.3, 61.2, 14.9; LCMS: m/z 399.1 [M + H]+; HRMS (ESI-TOF) m/z: [M + H]⁺ Calcd for C₁₇H₂₀FN₂O6_S 399.1026; Found 399.1016; HPLC: t _R 6.37 min, > 92% purity at 254 nm.

N′-(5-Ethoxy-2-fluoro-3-methylbenzoyl)-1H-indole-6-sulfonohydrazide (10c)

White amorphous solid (19%); ¹H NMR (400 MHz, DMSO-d ₆) δ 11.61 (s, 1H), 10.41 (d, J = 3.6 Hz, 1H), 9.78 (d, J = 3.6 Hz, 1H), 7.95 (s, 1H), 7.67 (d, J = 8.4 Hz, 1H), 7.63 (t, J = 2.8 Hz, 1H), 7.48 (dd, J = 8.4, 1.6 Hz, 1H), 6.95 (dd, J = 5.6, 3.1 Hz, 1H), 6.58 – 6.54 (m, 2H), 3.95 (q, J = 6.9 Hz, 2H), 2.19 (d, J = 1.7 Hz, 3H), 1.30 (t, J = 7.0 Hz, 3H); LCMS: m/z 391.9 [M + H]⁺; HRMS (ESI-TOF) m/z: [M + H]⁺ Calcd for C₁₈H₁₉FN₃O₄S 392.1080; Found 392.1081; HPLC: t _R 5.59 min, > 99% (254 nm).

N′-(5-Ethoxy-2-fluoro-3-methylbenzoyl)­benzo­[b]­thiophene-2-sulfonohydrazide (10d)

White amorphous solid (82%); H NMR (400 MHz, DMSO-d ₆) δ 10.62 (s, 1H), 10.49 (s, 1H), 8.11–8.03 (m, 3H), 7.54–7.47 (m, 2H), 6.98 (dd, J = 5.6, 3.1 Hz, 1H), 6.69 (dd, J = 4.7, 3.3 Hz, 1H), 3.98 (q, J = 7.0 Hz, 2H), 2.20 (d, J = 1.8 Hz, 3H), 1.30 (t, J = 7.0 Hz, 3H); ¹³C NMR (100 MHz, DMSO-d ₆) δ 163.7, 158.5, 154.2, 152.2, 142.1, 140.2, 138.1, 137.7, 130.8, 130.0, 127.7, 126.3, 125.8, 123.4, 120.5, 112.3, 64.2, 15.0; LCMS: m/z 408.9 [M + H]⁺; HRMS (ESI-TOF) m/z: [M + H]⁺ Calcd for C₁₈H₁₈FN₂O₄S₂ 409.0692; Found 409.0696; HPLC: t _R 6.19 min, > 95% purity at 254 nm.

3-Fluoro-N′-(3-fluoro-5-(pyridin-2-yl)­benzoyl)­thiophene-2-sulfonohydrazide (10e)

White amorphous solid (78%); ¹H NMR (400 MHz, DMSO-d ₆) δ 11.07 (d, J = 2.6 Hz, 1H), 10.59 (d, J = 2.7 Hz, 1H), 8.72 (ddd, J = 4.8, 1.7, 0.9 Hz, 1H), 8.36 (d, J = 1.3 Hz, 1H), 8.21–8.03 (m, 2H), 8.03 – 7.87 (m, 2H), 7.65–7.51 (m, 1H), 7.45 (ddd, J = 7.5, 4.8, 1.0 Hz, 1H), 7.12 (d, J = 5.6 Hz, 1H); ¹³C NMR (100 MHz, DMSO-d ₆) δ 164.7, 164.1, 161.6, 157.6, 154.2, 150.2, 142.0, 138.1, 134.9, 132.9, 124.2, 122.1, 121.3, 119.0, 117.2, 115.2; LCMS: m/z 395.8 [M + H]⁺; HRMS (ESI-TOF) m/z: [M + H]⁺ Calcd for C₁₆H₁₂F₂N₃O₃S₂ 396.0288; Found 396.0291; HPLC: t _R 4.32 min, > 99% purity at 254 nm.

5-Ethoxy-2-fluorobenzohydrazide (11b)

The ester was dissolved in EtOH to which hydrazine monohydrate (10 equiv) was added and the reaction mixture was allowed to reflux for 16 h. The resulting mixture was then cooled to room temperature and concentrated in vacuo. To the resulting mixture water was added and the resulting precipitate was filtered, washed with water and dried under vacuum to afford the compound of interest (67%). ¹H NMR (400 MHz, DMSO) δ 9.48 (s, 1H), 7.30–7.10 (m, 1H), 7.10–6.88 (m, 2H), 4.52 (bs, 2H), 4.02 (q, J = 7.0 Hz, 2H), 1.31 (t, J = 7.0 Hz, 3H); LCMS: m/z 199.0 [M + H]⁺.

Ethyl 5-ethoxy-2-fluorobenzoate (14)

To the solution of 2-fluoro-5-hydroxybenzoic acid (4.7 g, 30 mmol) in DMF (15 mL) was added iodoethane (6.1 mL, 75 mmol) and K₂CO₃ (12.4 g, 90 mmol). The resulting mixture was stirred at 50 °C for 17 h. Upon all starting material consumed, the reaction was cooled to room temperature and diluted with EtOAc (50 mL). The organic layer was washed with H₂O (120 mL) and brine, and then dried over MgSO₄. The filtrate was concentrated under reduced pressure to give the crude mixture. A further purification of crude mixture though silica gel flash chromatography (eluent: PE/EA = 6:1) provides the desired compound as a colorless oil (2.2 g, 35% yield). ¹H NMR (400 MHz, CDCl₃) δ 7.38 (ddd, J = 5.8, 2.6, 1.0 Hz, 1H), 7.04–7.00 (m, 2H), 4.39 (q, J = 7.1 Hz, 2H), 4.03 (q, J = 7.0 Hz, 2H), 1.40 (q, J = 7.2 Hz, 6H).

3-(2-Hydroxyethoxy)­benzenesulfonic acid (16)

A stirred mixture of phenol (1.0 equiv), ethylene carbonate (3.3 equiv) and K₂CO₃ (3.0) in DMF (0.1 M) was heated at 110 °C for 18 h. The reaction was allowed to cool and was filtered. The filtrate was concentrated in vacuo and the residue was taken up in brine and extracted with EtOAc (x3). The combined organic layers were dried with MgSO₄, filtered and concentrated in vacuo. The product was purified by flash chromatography (gradient elution with 0–30% MeOH/EtOAc) to yield desired compound (62%). ¹H NMR (400 MHz, DMSO) δ 7.30–7.07 (m, 3H), 6.91–6.82 (m, 1H), 3.96 (t, J = 5.0 Hz, 2H), 3.70 (t J = 5.3 Hz, 2H); LCMS: m/z 217.0 [M-H]⁺.

3-(2-Hydroxyethoxy)­benzenesulfonyl chloride (17)

To a stirred solution of 16 (1 equiv) in dry DCM (0.1 M) was added SOCl₂ (1.01 equiv) at room temperature and the resulting mixture was allowed to reflux for 16 h. The mixture was cooled to room temperature and all volatiles were evaporated under reduced pressure. The resulting crude product was used directly in the next step without any further analysis.

N-(2-(Oxazol-2-yl)-2-phenylethyl)-2H-thieno­[2,3-e]­[1,2,4]­thiadiazine-3-carboxamide 1,1 dioxide (18a)

To a solution of ethyl 2H-thieno­[2,3-e]­[1,2,4]­thiadiazine-3-carboxylate 1,1-dioxide (26.0 mg, 0.1 mmol) and 4Å molecular sieves (50 mg) in xylene (2 mL) was added 2-(oxazol-2-yl)-2-phenylethan-1-amine (28.3 mg, 0.15 mmol). The reaction was stirred at 140 °C for 12 h. Upon completion, the reaction was diluted with EtOAc and filtered via a Celite pad. The filtrate was concentrated and purified by silica gel flash chromatography to deliver the title product as a white solid (5.1 mg, 12% yield). ¹H NMR (400 MHz, DMSO-d ₆) δ 8.03 (s, 1H), 7.33 (d, J = 6.9 Hz, 2H), 7.28 (d, J = 7.1 Hz, 3H), 7.20 (d, J = 3.7 Hz, 1H), 7.16 (s, 1H), 7.08 (d, J = 7.0 Hz, 2H), 4.61 (t, J = 7.5 Hz, 1H), 3.91 (dd, J = 7.8, 5.7 Hz, 1H), 3.82 (q, J = 6.7 Hz, 1H); MS (m/z) 402.9 [M + H]⁺.

6-Chloro-N-(2-(oxazol-2-yl)-2-phenylethyl)-2H-thieno­[2,3-e]­[1,2,4]­thiadiazine-3-carboxamide 1,1-dioxide (18b)

To a solution of ethyl 7-chloro-2H-benzo­[e]­[1,2,4]­thiadiazine-3-carboxylate 1,1-dioxide (57.7 mg, 0.2 mmol) and 4Å molecular sieves (50 mg) in xylene (2 mL) was added 2-(oxazol-2-yl)-2-phenylethan-1-amine (56.5 mg, 0.3 mmol). The reaction was stirred at 140 °C for 12 h. Upon completion, the reaction was diluted with EtOAc and filtered via a Celite pad. The filtrate was concentrated and purified by silica gel flash chromatography to deliver the title product as a white solid (30.1 mg, 35% yield); ¹H NMR (400 MHz, DMSO-d ₆) δ 8.96 (s, 1H), 8.01 (s, 1H), 7.36–7.30 (m, 3H), 7.29–7.24 (m, 3H), 7.18 (s, 1H), 4.62 (t, J = 7.5 Hz, 1H), 3.99–3.91 (m, 1H), 3.86–3.79 (m, 1H); ¹³C NMR (101 MHz, DMSO-d ₆) δ 163.52, 156.72, 146.41, 143.65, 140.01, 138.43, 137.99, 128.79, 128.01, 127.57, 127.05, 119.04, 117.45, 43.76, 43.09; MS (m/z) 437.0 [M + H]⁺.

N-(2-(Oxazol-2-yl)-2-phenylethyl)-2H-thieno­[3,2-e]­[1,2,4]­thiadiazine-3-carboxamide 1,1-dioxide (18c)

To a solution of 2-(oxazol-2-yl)-2-phenylethan-1-amine (5 mg, 0.03 mmol) in EtOH (0.1 mL) was added ethyl 2H-thieno­[3,2-e]­[1,2,4]­thiadiazine-3-carboxylate 1,1-dioxide (7 mg, 0.02 mmol). The reaction was irradiated in a microwave at 120 °C for 2 h. The reaction was cooled and the precipitate was filtered. The solid was washed with EtOH (2 mL) and air-dried to give the title compound (5 mg, 50% yield) as a white solid; ¹H NMR (400 MHz, DMSO-d ₆) δ 9.24 (s, 1H), 8.07 (d, J = 5.3 Hz, 1H), 8.04 (d, J = 0.8 Hz, 1H), 7.39–7.14 (m, 6H), 4.66 (t, J = 7.6 Hz, 1H), 4.00 (ddd, J = 13.2, 7.5, 5.8 Hz, 1H), 3.92–3.77 (m, 1H); MS (m/z) 402.9 [M + H]⁺.

6-Chloro-N-(2-(oxazol-2-yl)-2-phenylethyl)-2H-thieno­[3,2-e]­[1,2,4]­thiadiazine-3-carboxamide 1,1-dioxide (18d)

To a solution of 2-(oxazol-2-yl)-2-phenylethan-1-amine (5 mg, 0.03 mmol) in EtOH (0.1 mL) was added ethyl 6-chloro-2H-thieno­[3,2-e]­[1,2,4]­thiadiazine-3-carboxylate 1,1-dioxide (7 mg, 0.02 mmol). The reaction was irradiated in a microwave at 120 °C for 2 h. The reaction was cooled and the precipitate filtered. The solid was washed with EtOH (2 mL) and air-dried to give the title compound (5 mg, 50% yield) as a white solid; ¹H NMR (400 MHz, CD₃OD) δ 7.88 (d, J = 0.9 Hz, 1H), 7.41–7.28 (m, 5H), 7.19 (d, J = 0.8 Hz, 1H), 7.13 (s, 1H), 4.66–4.60 (m, 1H), 4.08 (dd, J = 13.4, 8.2 Hz, 1H), 3.98 (dd, J = 7.0, 13.4 Hz, 1H); MS (m/z) 434.9 [M + H]⁺.

2,5-Dichlorothiophene-3-sulfonamide (20b)

To a solution of 2,5-dichlorothiophene-3-sulfonyl chloride (1.0 g, 4.0 mmol) in ethanol (5 mL) was add ammonia solution (2 M in EtOH, 5 mL, 10 mmol) at room temperature. The reaction was stirred at room temperature for 2 h. Upon completion, the reaction was diluted with water (20 mL) and extracted with EtOAc (10 mL x 3). The combined organic phase was washed with brine, dried over Na₂SO₄, and concentrated to give the crude solid. The crude solid was washed with diethyl ether to give the title compound as a white solid (0.9 g, 96% yield). ¹H NMR (400 MHz, DMSO-d ₆) δ 7.81 (s, 2H), 7.31 (d, J = 1.6 Hz, 1H).¹³C NMR (100 MHz, DMSO-d ₆) δ 140.6, 127.4, 126.9, 126.3.

Ethyl 2H-thieno­[2,3-e]­[1,2,4]­thiadiazine-3-carboxylate 1,1-dioxide (21a)

To an oven-dried round-bottom flash was charged with 2-chlorothiophene-3-sulfonamide (98.8 mg, 0.5 mmol), ethyl 2-ethoxy-2-iminoacetate (87.1 mg, 0.6 mmol), Cs₂CO₃ (325.8 mg, 1.0 mmol) and CuI (7.2 mmol, 0.05 mmol). The reaction mixture was purged with nitrogen and DMF (2 mL) was added at room temperature. The reaction was heated to 80 °C for 12 h. Upon completion, reaction was quenched by adding water and extracted with EtOAc (5 mL x 3). The combined organic phase was washed with brine, dried over Na₂SO₄, and concentrated. The crude mixture was purified by silica gel flash chromatography to provide the title product as a white solid (32.5 mg, 25% yield). ¹H NMR (400 MHz, CDCl₃) δ 7.05 (s, 1H), 6.62 (s, 1H), 4.42–4.34 (m, 2H), 1.40 (td, J = 7.2, 1.3 Hz, 3H).

Ethyl 6-chloro-2H-thieno­[2,3-e]­[1,2,4]­thiadiazine-3-carboxylate 1,1-dioxide (21b)

To an oven-dried round-bottom flask was charged with 2,5-dichlorothiophene-3-sulfonamide (116.0 mg, 0.5 mmol), ethyl 2-ethoxy-2-iminoacetate (87.1 mg, 0.6 mmol), Cs₂CO₃ (325.8 mg, 1.0 mmol) and CuI (7.2 mmol, 0.05 mmol). The reaction mixture was purged with nitrogen and DMF (2 mL) was added at room temperature. The reaction was heated to 80 °C for 12 h. Upon completion, the reaction was quenched by adding water and extracted with EtOAc (5 mL x 3). The combined organic phase was washed with brine, dried over Na₂SO₄, and concentrated. The crude mixture was purified by silica gel flash chromatography to provide the title product as a white solid (43.1 mg, 30% yield). ¹H NMR (400 MHz, DMSO-d ₆) δ 7.53 (d, J = 2.1 Hz, 1H), 4.54 – 4.33 (m, 2H), 1.34 (td, J = 7.2, 2.2 Hz, 3H); ¹³C NMR (100 MHz, DMSO-d ₆) δ 159.5, 143.0, 141.4, 125.8, 119.6, 118.0, 64.0, 14.3; MS (m/z) 292.9 [M-H]⁻.

N-(tert-Butyl)-5-chlorothiophene-2-sulfonamide (23b) ,

A solution of tert-butylamine (0.15 g, 2.1 mmol) in THF (2 mL/mmol) was cooled 0 °C and then a solution of the 5-chlorothiophene-2-sulfonyl chloride (0.43 g, 2.0 mmol) in THF (4 mL) was added over a period of 5 min. Then add subsequently Et₃N (0.24 g, 2.4 mmol) at the same temperature. Allow the reaction mixture to warm to room temperature and stir for 24 h and water (30 mL) was added. The organic phase was isolated, washed with brine (2 × 20 mL) and then dried over MgSO4 and reduced under vacuum. The oily mixture was purified by flash chromatography (PE/EtOAc = 10:1) to afford the tittle compound as a pale-yellow solid (0.4 g, 79% yield); ¹H NMR (400 MHz, CDCl₃) δ 7.38 (d, J = 4.0 Hz, 1H), 6.87 (d, J = 4.0 Hz, 1H), 4.62 (s, 1H), 1.31 (s, 9H); MS (m/z) 252.0 [M + H]⁺.

3-Amino-N-(tert-butyl)­thiophene-2-sulfonamide (24a)

i) A solution of N-(tert-butyl)­thiophene-2-sulfonamide (0.44 g, 2.0 mmol) in dry THF (10 mL) was cooled to −70 °C, and n-BuLi (1 M in hexane, 4 mL, 4.0 mmol) was added, maintaining the temperature < −65 °C. After this was added, the mixture was allowed to warm to −20 °C, stirred at this temperature for 30 min. ii) A solution of p-toluenesulfonyl azide (0.45 g, 2.3 mmol) in dry THF (2 mL) was added, maintaining the temperature at −20 °C, and the cooling bath was removed. After the mixture had reached room temperature, water (20 mL) was added. The organic phase was isolated, and the aqueous phase was extracted with toluene (3 × 15 mL). iii) To the combined organic phases was added hexadecyltributylphosphonium bromide (0.1 g, 0.2 mmol) followed by the dropwise addition of a solution of sodium borohydride (0.9 g, 2.4 mmol) in MeOH (5 mL) with stirring and cooling to room temperature. The mixture was stirred overnight at room temperature, and water (30 mL) was added. The organic phase was isolated, washed with water (2 × 30 mL), dried, and evaporated to dryness. The oily residue was dissolved in EtOAc (20 mL) and washed with 1 N NaOH (6 × 20 mL). The organic phase was dried with Na₂SO₄ and evaporated under vacuum. The oily mixture was purified by flash chromatography (PE/EtOAc = 4:1) to afford the title compound as a pale-yellow oil (140 mg, 30% yield); ¹H NMR (400 MHz, CDCl₃) δ 7.32 (d, J = 5.4 Hz, 1H), 6.53 (d, J = 5.4 Hz, 1H), 1.28 (s, 9H); MS (m/z) 235.0 [M + H]⁺.

3-Aminothiophene-2-sulfonamide HCl Salt (24a)

3-amino-N-(tert-butyl)­thiophene-2-sulfonamide (130 mg, 0.55 mol) was heated with stirring at 50 °C in concentrated HCl (2 mL) for 2.5 h. The reaction mixture was evaporated to give a brown solid and converted for the next step without further purification.

3-Amino-N-(tert-butyl)-5-chlorothiophene-2-sulfonamide (24b) ,

i) A solution of N-(tert-butyl)-5-chlorothiophene-2-sulfonamide (0.5 g, 2.0 mmol) in dry THF (5 mL) was cooled to −70 °C, and n-BuLi (1 M in hexane, 4 mL, 4.0 mmol) was added, maintaining the temperature < −65 °C. After this was added, the mixture was allowed to warm to −20 °C, stirred at this temperature for 30 min. ii) A solution of p-toluenesulfonyl azide (0.45 g, 2.3 mmol) in dry THF (5 mL) was added, maintaining the temperature at −20 °C, and the cooling bath was removed. After the mixture had reached room temperature, water (20 mL) was added. The organic phase was isolated, and the aqueous phase was extracted with toluene (3 × 15 mL). iii) To the combined organic phases was added hexadecyltributylphosphonium bromide (0.1 g, 0.2 mmol) followed by the dropwise addition of a solution of sodium borohydride (0.9 g, 2.4 mmol) in MeOH (5 mL) with stirring and cooling to room temperature. The mixture was stirred overnight at room temperature, and water (30 mL) was added. The organic phase was isolated, washed with water (2 × 30 mL), dried, and evaporated to dryness. The oily residue was dissolved in EtOAc (20 mL) and washed with 1 N NaOH (6 × 20 mL). The organic phase was dried with Na₂SO₄ and evaporated under vacuum. The oily mixture was purified by flash chromatography (PE/EtOAc = 4:1) to afford the tittle compound as a pale-yellow oil (0.3 g, 61% yield); ¹H NMR (400 MHz, CDCl₃) δ 6.36 (s, 1H), 1.30 (s, 9H); MS (m/z) 269.0 [M + H]⁺.

3-Amino-5-chlorothiophene-2-sulfonamide HCl Salt (24b)

3-amino-N-(tert-butyl)-5-chlorothiophene-2-sulfonamide (0.27g, 1 mmol) was heated with stirring at 50 °C in concentrated HCl (2 mL) for 2.5 h. The reaction mixture was evaporated to give a brown solid, and converted for the next step without further purification.

Ethyl 2H-thieno­[3,2-e]­[1,2,4]­thiadiazine-3-carboxylate 1,1-dioxide (25a)

To a solution of 3-aminothiophene-2-sulfonamide HCl Salt (130 mg, 0.55 mol) in AcOH (10 mL/mmol) was added ethyl carbonocyanidate (0.54 g, 5.5 mmol) and the mixture was stirred at RT under N₂ for 5 min. Concentrated aqueous HCI (Conc.HCl: AcOH = 1:40, 1 mL) was then added and the mixture was heated at 85 °C for 4 h. The mixture was concentrated under reduced pressure to dryness and was purified by flash chromatography (PE/EtOAc, 1:1) to afford the tittle compound as a white solid (30 mg, 21% yield); ¹H NMR (400 MHz, Methanol-d ₄) δ 7.96 (d, J = 5.4 Hz, 1H), 7.16 (d, J = 5.4 Hz, 1H), 4.47 (q, J = 7.2 Hz, 2H), 1.43 (t, J = 7.1 Hz, 3H); MS (m/z) 261.0 [M + H]⁺.

Ethyl 6-chloro-2H-thieno­[3,2-e]­[1,2,4]­thiadiazine-3-carboxylate 1,1-dioxide (25b)

To a solution of 3-amino-5-chlorothiophene-2-sulfonamide HCl salt (0.25 g, 1 mmol) in AcOH (10 mL) was added ethyl carbonocyanidate (1 g, 10 mmol) and the mixture was stirred at RT under N₂ for 5 min. Concentrated aqueous HCI (Conc.HCl: AcOH = 1:40, 2 mL) was then added and the mixture was heated at 85 °C for 4 h. The mixture was concentrated under reduced pressure to dryness and was purified by flash chromatography (PE/EtOAc = 1:1) to afford the tittle compound as a white solid (73.5 mg, 25% yield); ¹H NMR (400 MHz, Methanol-d ₄) δ 7.10 (s, 1H), 4.46 (q, J = 7.1 Hz, 2H), 1.43 (t, J = 7.1 Hz, 3H); MS (m/z) 293.0 [M + H]⁺.

KAT6A AlphaScreen Biochemical Acetylation Competition Assay

Compounds were tested for inhibitory activity against KAT6A using a previously reported assay format.

Revised KAT6A AlphaScreen Biochemical Acetylation Competition Assay

AlphaScreen technology (PerkinElmer) was utilized to assess the efficacy of small molecule inhibitors of MOZ-HAT activity in an 11-point dose curve 384-well plate format, as similarly described in Falk et al., with the following modifications.

Reactions (8 μL per well) were prepared in AlphaPlate-384 shallow well plates (Revvity, #6008350) with a final concentration of 10 μM AcCoA (except for 4, which was 15 μM) (Merck, #10101893001), 0.1 μg/mL anti-Histone H4 (acetyl K8) antibody (Abcam, #ab15823), 50 nM pHis₆-hMOZ with human MOZ MYST domain (residues 497–780) (a gift from Dr. Stefan Hermans, St Vincent’s Institute Medical Research, Melbourne, Australia), 100 nM biotinylated N-terminal histone H4 peptide (SGRGKG­GKGLGK­GGAKRH­RKVGGK-biotin, Merck, #12–405) prepared in assay buffer (100 mM Tris-HCl pH 7.8, 15 mM NaCl, 1 mM EDTA, 0.01% Tween-20, 1.5 mM DTT, and 0.5% BSA). Unless otherwise described, all assay agents were diluted in assay buffer prior to use. Before commencing the assay, compounds were freshly dissolved in DMSO, and an 11-point 3.5x serially diluted dose curve between 25 mM and 25.9 μM for each compound was then prepared in DMSO in a compound source plate (PerkinElmer, #6008590) using the JANUS G3 varispan arm (PerkinElmer). Next, an intermediate plate (PerkinElmer, #6008590) containing concentrated compound (4× concentrated) was prepared using assay buffer, to achieve a final dose range in the assay of between 125 μM and 0.129 nM. Within each assay plate, DMSO only (0.33% final concentration) and AcCoA substrate omitted controls were included. In addition, we also included compounds previously identified to inhibit KAT6A activity (Falk et al., PMID: 22086275) in each plate as further controls. To ensure assay robustness, for all assays, we utilized Z’ factor calculations to assess the dynamic range of the assay signal; all assays conducted achieved a Z’ factor >0.6, indicating an excellent dynamic range between positive and negative controls.

To conduct the assays, 4 μL/well of AcCoA substrate and anti-Histone H4 (acetyl K8) antibody was first dispensed into each assay plate using a Multidrop Combi Dispenser (Thermo Fisher Scientific). Then, 2 μL/well of the 4x concentrated serially diluted compounds were transferred from the intermediate compound plate to duplicate assay plates using the JANUS G3 Automated Workstation Modular Dispense Technology (MDT)-384 Tip Head (PerkinElmer). Next, 2 μL/well of the pHis₆-hMOZ and biotinylated N-terminal histone H4 peptide mix was dispensed to the assay plate with Multidrop Combi Dispenser to initiate the enzymatic reaction. Plates were sealed, shaken for 10 s with Multidrop Combi Dispenser plate shaker (600 rpm), centrifuged at 500×g for 1 min, and then incubated at room temperature for 90 min.

Following these steps, 8 μg/mL (final concentration) of both AlphaScreen Protein A acceptor beads and streptavidin donor beads (Revvity, #6760617c; prepared in assay buffer, dispensed a total volume of 4 μL/well) was added to the reactions in assay plates using the Multidrop Combi Dispenser. The assay plates were then resealed, centrifuged at 500 x g for 1 min, shaken for 10 s with Multidrop Combi Dispenser plate shaker (600 rpm), and then incubated at room temperature for a further 2 h in low light conditions. After incubation, the plates were read using EnVision Multimode microplate reader equipped with an AlphaScreen module (Revvity) utilizing the default factory AlphaScreen settings. Data was then analyzed, where IC₅₀ values with a 95% confidence interval (CI) were determined by fitting the percent inhibition of AlphaScreen signal versus compound concentration (11 dilutions, duplicate plates) to a log (inhibitor) vs response – variable slope, four-parameter logistic model using GraphPad Prism 9.2.0 software.

$$\%\mathrm{i}\mathrm{n}\mathrm{h}\mathrm{i}\mathrm{b}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}=(1-\frac{x-\mu (\mathrm{n}\mathrm{e}\mathrm{g})}{\mu (\mathrm{p}\mathrm{o}\mathrm{s})-\mu (\mathrm{n}\mathrm{e}\mathrm{g})})\times 100$$

x is sample AlphaScreen signal. μ is mean AlphaScreen signal. Positive control: no compound (DMSO), n = 16. Negative control: AcCoA omitted, n = 16.

Cell Proliferation Assay

The cell-based proliferation assay was done by Reaction Biology Corporation (Freiburg, Germany). Two breast cancer cell lines were used: ZR-75–1, which was cultured in DMEM supplemented with 10% fetal calf serum (FCS), and T-47D, which was cultured in RPMI-1640 supplemented with 10% FCS and 0.2 U/mL insulin. For the assays, cells were seeded in white cell culture-treated flat and clear bottom 384-well plates and incubated at 37 °C overnight before adding compounds. After incubation for 7 days at 37 °C at 5% CO₂, cell plates were equilibrated to room temperature for one hour, CellTiterGlo reagent (Promega) was added and luminescence was measured approximately an hour later using a luminometer (EnVision, PerkinElmer).

The compound treatment of cells started 1 day after seeding with a final DMSO concentration of 0.1% and was performed by nanodrop-dispensing using a Tecan Dispenser (D300e). 0.1% DMSO (solvent) served as high control (100% viability) and CTx-648 (PF-9363; 6) (1.0 × 10^–5 M) as low control (0% viability). Compounds were tested at the highest concentrations of 1.0 × 10^–5 M with 1:5 dilution up to 5.12 × 10^–12 M (10 series of concentrations used in the assay).

The raw data were converted into percent cell viability relative to the high and low control, which were set to 100% and 0%, respectively. The IC₅₀ calculation was performed using GraphPad Prism software with a variable slope sigmoidal response fitting model using 100% viability as the top constraint.

Structural Biology

Crystallization of MYST^cryst in Complex with Inhibitors

The inhibitor complexed crystals of MYST^cryst were all obtained by direct co-crystallization. The highest quality crystals grew in hanging drops of 1 μL protein solution (1.9–5.6 mg/mL MYST^cryst, 50 mM HEPES pH 7.5, 200 mM NaCl, 0.2–0.4 mM inhibitor, 2% v/v DMSO) and 1 μL precipitant solution (for 8: 0.1 M Tris pH 7.0, 0.25 M MgCl₂, 8% PEG8K; for 6: 0.25 M K–Na Tartrate, 20% PEG3350; for 7: 0.1 M Bis-Tris HCl pH 5.5., 0.2 M LiSO₄, 9% PEG3350; for 10a–e and 18c: 0.1 M HEPES pH 7.5, 0.2–0.3 M NaCl, 12.5–15% PEG3350). Crystals were mounted in cryoloops, transferred to a cryoprotectant solution supplemented with 20% glycerol, and flash-cooled by rapid immersion in liquid nitrogen.

Data Collection and Processing

The X-ray diffraction data for MYST^cryst in complex with inhibitor 6 were collected in-house using a Rigaku XtaLAB Synergy-S single crystal X-ray diffractometer and integrated using CrysAlis PRO software (Agilent). All other X-ray diffraction data were collected at the MX1 and MX2 beamlines of the Australian Synchrotron (Clayton, Victoria, Australia) and integrated using XDS. All data were scaled with AIMLESS of the CCP4 program suite, labeling 5% of the reflections for the Rfree set. Data collection statistics are shown in Supporting Information 2.

Structure Determination and Refinement

All structures were determined by molecular replacement using the structure of MYST^cryst in complex with inhibitor WM-8014 (PDB ID: 6BA2) as the search model with PHASER. All structures were refined through cycles of manual building using COOT and refinement using PHENIX Refine of the PHENIX program suite. A restraint library for each inhibitor was developed through PHENIX eLBOW. Structure validation was monitored with MolProbity and refinement statistics are shown in Supporting Information 2. The Ramachandran plot for the inhibitor 10d complex structure showed one outlier, Ala759, in an exposed loop with poorly defined electron density. The Ramachandran plot for all other structures showed no outliers.

Molecular Modeling, Generation of Protein Domains and Sites Diagram, and Physicochemical Properties Calculations

All molecular structures were drawn using ChemDraw (version 21.0.0). ICM-Pro (version 3.9–4) was used for protein visualization, analysis of protein–ligand interactions, and protein sequence alignment. The domain and site diagram of MYST KATs was generated using DOG (version 2.0). HBA, HBD, pK _a, logP, and logD values were predicted using the Percepta Platform (ACD/Lab). tPSA values were calculated with ChemDraw (version 21.0.0).

Protein Data Bank (PDB) accession codes are assigned in parentheses as follows: 6 (9OO9), 7 (9OOA), 8 (9OOB), 10a (9OOC), 10b (9OOH), 10c (9OOD), 10d (9OOE), 10e (9OOF), 18c (9OOJ). The authors will release the atomic coordinates and experimental data upon article publication.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jmedchem.5c01426.

• MOZ-HAT biochemical assay setup, benchmark compounds fused [1,2,4]­thiadiazines 4, benzisoxazole sulfonamides 6 and 7, acyl sulfonamide 8 and the N-(2-oxoethyl) sulfanilamide derived 9 resyntheses, and acylsulfonohydrazide pK _a measurement (PDF)

• Co-crystal structure collection and refinement statistics (XLSX)

• Molecular Formula Strings (CSV)

▲.

Manas AI, Inc. New York, New York 10011, United States

¶.

A.S. and J.J. contributed equally. This manuscript was written through the contributions of all authors. All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.