Identification of 5-Thiocyanatothiazol-2-amines Disrupting WDR5-MYC Protein–Protein Interactions

Abstract

MYC amplification is frequently observed in approximately 50% of human cancers, rendering it a highly desired anticancer target. Given the challenge of direct pharmacological inhibiting of MYC, impairing the interaction of MYC and its key cofactor WDR5 has been proposed as a promising strategy for MYC-driven cancer treatment. Herein, we report the discovery of 5-thiocyanatothiazol-2-amines that disrupt the WDR5-MYC interaction. Hit fragments were initially identified in a fluorescence polarization (FP)-based screening of an in-house library, and structural–activity relationship exploration resulted in the lead compounds 4m and 4o with potent inhibitory activities on WDR5-MYC interaction (K_i = 2.4 μM for 4m; K_i = 1.0 μM for 4o). These compounds were further validated via differential scanning fluorimetry (DSF) and coimmunoprecipitation (Co-IP). Moreover, 4m and 4o exhibited good cellular activities with the IC₅₀ values at the micromolar level (IC₅₀ = 0.71–7.40 μM) against multiple MYC-driven cancer cell lines. Our findings afforded a potential small molecule blocking the WDR5-MYC interaction.

MYC protein is a family of transcription factor proteins, which includes C-MYC, L-MYC, and N-MYC.¹ MYC proteins regulate a vast array of cellular processes, including cell cycle, cell growth, metabolism, differentiation, apoptosis, transformation, and angiogenesis.² The dysregulation of MYC would cause extensive transcriptional reprogramming that drives increased cell proliferation, leading to tumor initiation and progression.³⁻⁵ However, modulation of MYC protein remains technologically challenging, mostly due to the lack of defined pockets for accommodating small molecular inhibitors in MYC protein.^6,7 Although several strategies have been investigated through blocking the protein–protein interactions (PPI) between MYC and other critical cofactors,^8,9 such as blocking MAX-MYC dimerization, MAX/MYC dimer-DNA interaction and MYC-RRAP interaction, most of these PPI inhibitors have not been advanced into further development (Figure 1).

Figure 1.

Representative small molecules inhibiting the interaction of MYC and its key cofactor protein.

Recent studies revealed that WD repeat-containing protein 5 (WDR5) serves as a crucial scaffolding protein in facilitating recruitment of MYC to chromatin, which is essential for tumor maintenance.¹⁰ The “WDR5-binding motif” (WBM) site of WDR5 binds to the highly conserved “MYC box” (MbIIIb) in the central domain of MYC, and binds to chromatin through colocalization with MYC to regulate protein synthesis and the downstream gene transcription.¹¹ A cell line carrying a switchable c-myc allele is shown defective for interaction with WDR5, and this mutant caused a rapid and complete tumor regression in vivo.¹² These findings have provided important clues that disrupting the WDR5-MYC interaction would be a focal point for effective MYC-driven cancer therapy. Although a few small molecular inhibitors were discovered targeting WDR5-MYC interaction via high-through screening recently (Figure 1),¹³⁻¹⁶ it is also necessary to provide more chemical space diversity for the tool boxes disrupting the WDR5-MYC interaction.

Herein, we conducted a fluorescence polarization (FP)-based screening of our in-house library (∼1000 compounds), and a distinct structure of 5-thiocyanatothiazol-2-amines was identified with good inhibitory activities. A structure–activity relationship study resulted in lead compounds 4m and 4o, and the capability of blocking the WDR5-MYC interaction was also evidenced by differential scanning fluorimetry (DSF) and co-immunoprecipitation (Co-IP). Moreover, 4m and 4o exhibited potent antiproliferative inhibitory activities against multiple MYC-driven cell lines. This forms a strong basis for the development of cellularly active small molecular inhibitors that disrupt WDR5-MYC interactions.

To identify the new chemotypes disrupting the WDR5-MYC interaction, we started with conducting an in-house drug-like library screening against recombinant constructs of WDR5 (1–334 aa) protein. A fluorescence polarization (FP) competitive assay was developed. The binding surface of MYC with WDR5 contains multiple basic residues such as “EEIDVV”, a motif conserved in all MYC proteins.^17,18 Thus, a synthetic 5-FAM-labeled peptide containing 9 amino acid residues (5FAM-PEG4-EEEIDVVSV) serves as a reporter probe. The sequence of the reporter probe is derived from the MYC MbIIIb protein sequence (residues E261 to V267) bound to WDR5 (Figure 2-a). The binding affinity of the small molecules is calculated by monitoring the changes in the FP signal upon varying the concentrations of compounds and fitting the data to a hyperbolic function, as described in the experimental section. In this FP-based assay, free MYC peptide showed a K_d value of 1.0 μM as the assay quality control (Figure 2-b). An in-house library with ∼1000 structurally diverse compounds was initially evaluated at a final concentration of 50 μM. This generated a 5-thiocyanatothiazol-2-amine 4a as an initial hit with a K_i value of 6.1 μM using DMSO as the negative control (Figure 2-c).

Figure 2.

In vitro profile of WDR5-MYC interaction inhibitor 4a identified from the FP screen. a) Illustration of the FP (fluorescence polarization) assay for MYC peptide and WDR5 interaction. b) Inhibitory curve of free MYC peptide in the FP assay. c) Inhibitory curve of 4a in the FP assay.

Encouraged by the results of the initial hit 4a, we further evaluated the analogues of 4a in our house library¹⁹ or synthesized 4a derivatives for revealing key structural features on the 5-thiocyanatothiazol-2-amine. For instance, the installation of an aryl group at the C-4 position of 4a proved advantageous to improve the activity (Table 1). The desired analogues 4b–4l and 4n (K_i = 0.9–4.1 μM, except 4g) exhibited better activities than 4a (K_i = 6.1 μM) in an FP assay. The 4-chlorophenyl (4b, K_i = 1.3 μM), 4-bromophenyl (4d, K_i = 0.9 μM), and 4-(methoxycarbonyl) phenyl (4n, K_i = 1.7 μM) substitutions turned out to be favorable on the C-4 position of 4a, leading to a 4–5-fold improvement in the K_i values over 4a (K_i = 6.1 μM). Except for 4b and 4c, it is noticed that the position of functional groups on the aryl ring had small impacts on the activity (4d–4f).

Table 1. First-Round SAR Study at the C-4 Position of the Hit Compound 4a.

^aK_i values were obtained by a fluorescence-polarization assay. K_i values represent the one of two or more independent, replicate determinations (CV < 0.3). The brackets indicate 95% confidence intervals for K_i values.

Based on the first-round SAR study, we then used the most biologically active compound 4b to explore the substitutions on the NH₂ group at the C-2 position (Table 1). To probe if the chemical diversity of the NH₂ group was essential for the activity, methyl (7a), pyridinyl (7b), amide (7c–7d), and sulfamide (7e) were synthesized. The installation of bulky aryl fragments (7b, 7d–7e) in the replacement of primary amine (4b) proved detrimental to the inhibition of the WDR5-MYC interaction, resulting in an almost complete loss in activity. To assess the importance of other functional groups on the NH₂ group at the C-2 position, we incorporated other alkyl groups through alkylation and amidation. These compounds retained their activities (7a: K_i = 1.2 μM; 7c: K_i = 1.6 μM). This indicated that only the substitutions with small steric effects on the NH₂ group can be tolerant.

Table 2. Second-Round SAR Study of NH₂ Group at the C-2 Position of 4b.

^aK_i values were obtained by a fluorescence-polarization assay. K_i values represent the one of two or more independent, replicate determinations (CV < 0.3). The brackets indicate 95% confidence intervals for K_i values.

In the second-round SAR study, it is revealed that the free amino group on the C-2 position of 4b is crucial for activity. Based on this, further modifications were carried out on the conserved amino group. But the role of thiocyano group (SCN) at the C-5 position of 4b is not clear yet. So, in the third-round SAR study, the replacement of SCN with other substitutions including hydrogen (H), halo (-Cl, -Br), polar group (−CN), and sulfur-containing groups (SCH₃, S(O)CH₃, S(O)₂CH₃) were all synthesized (Table 3). However, it was observed that any replacement of the SCN group resulted in inactive compounds (8–9, 11–13, and 16). The thiocyano group has been reported as a privileged covalent warhead that can probe an enzyme or protein binding site.²⁰ To understand if the SCN group serves as a potential covalent warhead in this study, time-dependent evaluation of 4b in the FP binding assay was performed according to the reported protocols^13,14 (Figure 3). 4b showed a time-dependent inhibition in the FP assay, and the biochemical activity was increased by ∼1.5-fold from 0.5 to 6 h along the maximal inhibitory improvement from 33.2% to 51.2%, while the reported WDR5-MYC inhibitor SADS-1 did not exhibit the time-dependent inhibition in the FP assay.

Table 3. Third-Round SAR Study of the C-5 Position of 4b.

compound	R3	FPA Inh %a
3b	H	0.9
8	Cl	–12
9	Br	–5
11	CN	1.1
12	SCH3	–21
13	S(O)CH3	0.4
16	S(O)2CH3	–10

^aFPA Inh %: Fluorescence polarization inhibition rate was tested with the targeted compound at a concentration of 50 μM and the reported values were represented as the mean (n = 2).

Figure 3.

Inhibition curves of compound 4b and SADS-1 in the FP assay with an indicated incubation time (from 30 to 360 min).

Given that the functional substitutions on the aryl group at the C-4 position did not cause a dramatic variation in the inhibition of the WDR5-MYC interaction, we further wonder whether to incorporate polar groups, such as the hydroxyl group (4m) and amide (4o) for tuning the physical-chemical characters of compounds. It was observed that both 4m and 4o displayed acceptable calculated logP values (4m and 4o: clogP = 1.79 and 1.74 vs 4b: clogP = 2.88) based on the ADMETlab (https://admetmesh.scbdd.com/service/evaluation/cal). In the FP assay, it was observed that amide analogue 4o (K_i = 1.0 μM) exhibited a similar activity as 4b (K_i = 1.3 μM), while the hydroxyl group analogue 4m (K_i = 2.4 μM) showed a slight drop in activity. But all the selected hits displayed more potent inhibition than SADS-1 (a reported WDR5/MYC interaction inhibitor) in our established FP assay¹³ (Table 4). Similar to 4b, covalent engagement of compound 4m and 4o was confirmed through a time-dependent FP inhibition (Figure 4a,b). The biochemical activities of 4m and 4o were increased by ∼1.4 to 2.5-fold from 0.5 to 6 h.

Table 4. Structural Optimization of the Hit Compound 4b.

^aK_i values were obtained by a fluorescence-polarization assay. K_i values represent the one of two or more independent, replicate determinations (CV < 0.3). The brackets indicate 95% confidence intervals for K_i values.

Figure 4.

(a, b) The inhibition curves of compound 4m and 4o in the FP assay with the indicated incubation time (from 30 to 360 min). (c) T_i identification curves of selected hits in the DSF. (d) CoIP assay to detect the interaction between His-tagged WDR5 protein and intracellular C-MYC-Flag after treatment with 4m or 4o.

In addition, the potencies of selected hits were further assayed by orthogonal methods. DSF was utilized to verify the binding of the compounds to the WDR5 protein. DDO2117 (a reported WDR5’s WIN site inhibitor)²¹ and SADS-1¹³ were chosen as the controls. The change in the inflection point temperature (T_i) of the WDR5 protein upon treatment with the compounds was monitored. Compared to the DMSO group, DDO2117 significantly increased the T_i value of WDR5 from 77.2 to 88.4 °C, a total increase of 11.2 °C. And the SADS-1 caused a slight increase on the T_i value of WDR5 with the ΔT_i value of +1.4 °C. When treated with 4b, 4m, or 4o, the T_i values of WDR5 were decreased in the range of −2.7 °C to −5.8 °C with the 4o inducing the highest thermal shift (Figure 4c). This suggests that the binding of 4b, 4m, or 4o will destabilize the WDR5 protein, potentially through a mechanism distinct from that of previously reported WDR5 inhibitors. Although the mechanism of compound-induced WDR5 destabilization is largely unknown, the thiocyano (SCN) group seems to be required. Next, we evaluated the effect of 5-thiocyanatothiazol-2-amine hits on WDR5-MYC interactions in the cellular context. We initially incubated the His-tagged WDR5 protein with 4m or 4o for 2 h. Subsequently, the protein was employed as bait to investigate its interaction with C-MYC overexpressed 293T cell lysate through Co-IP samples. As shown in Figure 4d, 4m and 4o could induce a clear dose-dependent disassociation between His-tagged WDR5 and the intracellular MYC proteins. Since we applied 1 μg of purified recombinant WDR5 protein (His-tagged) to precipitate its interesting cellular protein, the concentrations were much higher than those used in the in vitro assays.

We then conducted an antiproliferative activity study of 4m and 4o on cancer cell lines, which possessed MYC amplification. Cancer cell lines including the hematopoietic carcinoma cell line SUDHL4, RAMOS and RAJI, and the lung cancer cell line H1650 were selected. As shown in Table 5, the cell lines showed sensitivity to 5-thiocyanatothiazol-2-amines with the IC₅₀ values ranging from 0.71 to 7.40 μM. These promising anticancer cell growth activities of 5-thiocyanatothiazol-2-amine hits served as a basis for further development since most of the WDR5-MYC interaction inhibitors rarely reported their antitumor activities in a cellular assay.

Table 5. Antiproliferative Activities of Selected Leads against MYC-Driven Cancer Cell Lines.

		IC50/μMa
cancer lineage	cell lines	4m	4o
Hematopoietic (leukemia, lymphoma)	RAJI	0.71 ± 0.10	1.52 ± 0.53
	RAMOS	1.06 ± 0.14	1.51 ± 0.42
	SUDHL4	1.43 ± 0.24	2.20 ± 0.26
	HL-60	3.03 ± 0.77	5.32 ± 0.76
Solid tumor (lung cancer)	H1299	3.70 ± 0.10	7.40 ± 2.00

^aIC₅₀ values represent the mean ± SD of three independent experiments.

The synthetic route to 5-thiocyanatothiazol-2-amines analogues is depicted in Schemes 1, 2, and 3. The key intermediates 3b–3n were synthesized via cyclization from vinyl azides 1, α-bromoacetophenone 2, or acetophenone 5, respectively. The FeBr₃-catalyzed thiocyano group modification on the C-4 position of 2-aminothioazole ring affords the corresponding products 4b–4m according to our previous work.¹⁹ An amide coupling mediated by HATU of acid 3o with methylamine hydrochloride and subsequent thiocyano group incorporation afforded the desired product 4o. Then, the synthetic route depicted in Scheme 2 was utilized to investigate the SAR of the amino group at the C-2 position of the thioazole ring. In the case of alkyl substitution and pyridinyl substitution, the corresponding products (7a and 7b) were obtained from the cyclization of α-bromoacetophenone with substituted thiourea, following the SCN modification. In the case of amide substitution, the corresponding products (7c and 7d) were synthesized via amide coupling on the 4-(4-chlorophenyl) thiazol-2-amine 3b, following the thiocyano installation, while in the case of sulfamide substitution, the desired product 7e was synthesized via amide coupling on the existing 5-thiocyanothiazol-2-amine 3b with tosyl chloride and subsequent thiocyanate modification. Finally, the synthetic route depicted in Scheme 3 was followed to demonstrate the functional group replacement at the C-5 position of the 2-aminothiazole ring. The halo groups bearing products 8–9 were synthesized from the substituted reaction on the existing core of 3b. In the case of CN substitution, the synthesis route is quite different. The starting material methyl 4-chlorobenzoate was substituted by acetonitrile to obtain 10. Then a cyclization between 10 and thiourea affords 11. As for those with SCH₃, S(O)CH₃, and S(O)₂CH₃ substitution, targeted compounds were synthesized from 4′-chloroacetophenone. The desired 12 was obtained through the cyclization of α-brominated acetophenone di-Me sulfonium salt III·Br. Further oxidation of 12 generated corresponding products 13 and 16. 13 was directly obtained by oxidation of 12 with H₂O₂. 16 was generated from 12 in three steps. The C-2 amino group of 12 was first protected with Boc group and then oxidized by m-CPBA following by the N-Boc deprotection.

Scheme 1. Synthesis of 4a–4o.

Reagents and conditions: (a) KSCN, Pd(OAc)₂, CH₃CH₂CH₂OH, 80 °C, 12 h, 65–95%; (b) Thiourea, EtOH, reflux, 4 h, 49–93%; (c) FeBr₃, KSCN, MeCN, 80 °C, 12 h, 8–83%; (d) Br2, DCM, rt, 2 h, 90%; (e) 10% NaOH, 1 N HCl, rt, 6 h, 56%; (f) CH₃NH₂·HCl, DIPEA, HATU, DCM, rt, 6 h, 51%.

Scheme 2. Synthesis of 7a–7e.

Reagents and conditions: (a) R₂NHC(S)NH₂, EtOH, reflux, 4 h, 91%–97%; (b) R₂Cl, Et₃N, DMAP, DCM, ice bath to reflux, 42–71%; (c) FeBr₃, KSCN, MeCN, 80 °C, 12 h, 37–71%.

Scheme 3. Synthesis of 8–9, 11, 12–13, 16.

Reagents and conditions: (a) NCS, DMF, 0 °C, 2 h, 84%; (b) NBS, DMF, 0 °C, 2 h, 96%; (c) NaH, MeCN, THF, reflux, 5 h, 50%; (d) Thiourea, pyridine, I₂, EtOH, reflux, 5 h, 34%–72%; (e) HBr, DMSO, 60 °C, overnight, 86%; (f) thiourea, EtOH, reflux, 4 h, 75%; (g) H₂O₂, AcOH, rt, 2 h, 58%; (h) Boc₂O, Et₃N, DMAP, THF, 50 °C, overnight, 67%; (i) m-CPBA, MeOH, 50°C, 5 h, 50%; (j) CF₃COOH, DCM, rt, 2 h, 61%.

In summary, we disclose a well-characterized fragment 5-thiocyanatothiazol-2-amine with micromolar potency impairing the interaction between WDR5 and MYC. The potency of selected leads 4m and 4o was evidenced by multiple orthogonal methods, including FP, DSF, and Co-IP. Moreover, 4m and 4o displayed potent anticancer cellular activities against multiple MYC-driven cancer cell lines. The data provided in this work pave the way for the development of more potent molecules disrupting WDR5-MYC interaction for MYC-driven tumor therapy.

Glossary

Abbreviations

FP

fluorescence polarization

Co-IP

coimmunoprecipitation

DSF

differential scanning fluorimetry

PPI

protein–protein interactions

WDR5

WD repeat-containing protein 5

WBM

WDR5-binding motif

HATU

2-(7-azabenzotriazol-1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate

SAR

structure–activity relationship

m-CPBA

3-chloroperbenzoic acid

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsmedchemlett.4c00220.

• Experimental procedure and data, HPLC analysis of all final compounds (Table S1), HPLC traces and NMR spectra of intermediates and final biological tested compounds (PDF)

Author Contributions

^∇ H.W. and Y.Z. contributed equally.

The authors declare no competing financial interest.