Minimum structural requirements for inhibitors of the zinc finger protein TRAF6

Article abstract of DOI:10.1016/j.bmcl.2019.06.050

Zinc fingers have rarely been regarded as drug targets. On the contrary, the zinc-binding site of enzymes has often been considered a target of inhibitors. We previously developed a dithiol compound called SN-1 that binds to the zinc finger protein tumor necrosis factor receptor-associated factor 6 (TRAF6) and suppresses downstream nuclear factor-κB (NF-κB) signaling. To determine the minimal structure requirements of TRAF6 inhibitors based on SN-1, NF-κB inhibitory activity and cytotoxicity of its derivatives including new compounds were examined. SN-2, an oxidative type of prodrug of SN-1 with 2-nitrophenylthio groups via disulfide, has the minimum structure for an inhibitor of TRAF6, as seen with cellular experiments. The importance of two side chains with a thiol group was shown with molecular modelling. This study may lead to development of selective TRAF6 inhibitors in the near future.

Full text of DOI:10.1016/j.bmcl.2019.06.050

Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Minimum structural requirements for inhibitors of the zinc finger protein
TRAF6
Mohamed O. Radwana,b,c,1, Ryoko Kogaa,1, Tomohiro Hidaa, Tomohiko Ejimaa,
⁎
Yosuke Kanemarua, Hiroshi Tateishia, Yoshinari Okamotoa, Jun-ichiro Inoued, Mikako Fujitaa,
,
⁎
Masami Otsukaa,b,
a Medicinal and Biological Chemistry Science Farm Joint Research Laboratory, Faculty of Life Sciences, Kumamoto University, 5-1 Oe-honmachi, Chuo-ku, Kumamoto 862-
0973, Japan
b Department of Drug Discovery, Science Farm Ltd, 1-7-30-805 Kuhonji, Chuo-ku, Kumamoto 862-0976, Japan
c Chemistry of Natural Compounds Department, Pharmaceutical and Drug Industries Research Division, National Research Centre, Dokki 12622, Cairo, Egypt
d Division of Cellular and Molecular Biology, Institute of Medical Science, The University of Tokyo, 4-6-1 Shirokanedai, Minato-ku, Tokyo 108-8639, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
Zinc finger
TRAF6
Zinc fingers have rarely been regarded as drug targets. On the contrary, the zinc-binding site of enzymes has
often been considered a target of inhibitors. We previously developed a dithiol compound called SN-1 that binds
to the zinc finger protein tumor necrosis factor receptor-associated factor 6 (TRAF6) and suppresses downstream
nuclear factor-κB (NF-κB) signaling. To determine the minimal structure requirements of TRAF6 inhibitors based
on SN-1, NF-κB inhibitory activity and cytotoxicity of its derivatives including new compounds were examined.
SN-2, an oxidative type of prodrug of SN-1 with 2-nitrophenylthio groups via disulfide, has the minimum
structure for an inhibitor of TRAF6, as seen with cellular experiments. The importance of two side chains with a
thiol group was shown with molecular modelling. This study may lead to development of selective TRAF6
inhibitors in the near future.
NF-κB
Molecular docking
Introduction
substrates,12 and thus, this site has been considered a target of in-
hibitors. Among the inhibitors, captopril, an inhibitor of zinc-con-
Zinc is often present in the catalytic center of enzymes that catalyze
various biological reactions.1 In other cases, zinc contributes to protein
folding to maintain the stability of the protein,2–5 and to the functions
of the protein via formation of a zinc finger moiety.1 Initially, zinc
fingers were investigated as DNA binding motifs of transcription fac-
tors. Later, some cytoplasmic proteins were found to have zinc fingers
that exert these functions. One explicit example is tumor necrosis factor
receptor-associated factor 6 (TRAF6), a signaling transducer implicated
in cancer and inflammation.6,7 TRAF6 has five zinc fingers C-terminal
to the N-terminal RING finger domain, which has E3 ubiquitin ligase
activity against TRAF6 itself and other signaling proteins and leads to
downstream signaling.8–10 The first zinc finger regulates poly-ubiqui-
tination of TRAF6.11
taining angiotensin-converting enzyme,13 entered the market and is
used widely. On the other hand, zinc fingers have been rarely regarded
as drug targets. Previously, we synthesized a zinc protein-binding
compound called SN-114–17 (Fig. 1). SN-1 binds to the zinc finger pro-
tein, TRAF6, and suppresses its ubiquitination and downstream nuclear
factor-κB (NF-κB) signaling.18 Introduction of an SN-1 recognition
moiety that can specifically bind TRAF6 zinc fingers may eventually
lead to a selective TRAF6 inhibitor drug. In our endeavors to achieve
this goal, here we determined the minimal structure requirements of
TRAF6 inhibitors based on SN-1.
SN-1 has two thiol groups, and thus, the compound is easily oxi-
dized to cyclic disulfide oligomeric mixtures (SN-3) (Fig. 1). To sur-
mount this obstacle, we have used a compound with 2-nitrophenylthio
groups via disulfide (SN-2) as an oxidative type of prodrug of SN-1 in
The zinc-binding site of enzymes has a pocket to accommodate these
Abbreviations: DTT, DL-dithiothreitol; IL-1α, interleukin-1α; MTT, 3-(4,5-dimethylthial-2-yl)-2,5-diphenyltetrazalium bromide; NF-κB, nuclear factor-κB; NpSH, o-
nitrobenzenethiol; TRAF6, tumor necrosis factor receptor-associated factor 6
⁎ Corresponding authors at: Medicinal and Biological Chemistry Science Farm Joint Research Laboratory, Faculty of Life Sciences, Kumamoto University, 5-1 Oe-
honmachi, Chuo-ku, Kumamoto 862-0973, Japan.
E-mail addresses: mfujita@kumamoto-u.ac.jp (M. Fujita), motsuka@gpo.kumamoto-u.ac.jp (M. Otsuka).
1 These authors contributed equally to this work.
https://doi.org/10.1016/j.bmcl.2019.06.050
Received 17 May 2019; Received in revised form 15 June 2019; Accepted 26 June 2019
0960-894X/©2019ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:MohamedO.Radwan,etal.,Bioorganic&MedicinalChemistryLetters,https://doi.org/10.1016/j.bmcl.2019.06.050

M.O. Radwan, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Fig. 1. Structures of SN-1 and its derivatives used in this study. SN-114, SN-214, SN-316, Biotin-SN-117, Farnesyl-SN-121, and Dodecyl-SN-121 were previously reported.
SN-5, SN-6, SN-7, SN-8, and SN-9 were synthesized in this study. Bis(2-nitrophenyl)disulfide was commercially available.
cellular experiments.16,18 To explore the structure-activity relationship
in more detail, compounds SN-5 and SN-6 with phenyl and pyridinyl
groups, respectively, instead of the 4-dimethylaminopyridinyl group of
SN-2 were synthesized. Compound SN-9 that lacks one side chain of SN-
1, and SN-7 that does not have the dimethylamino group present in SN-
9 were also designed. Furthermore, compound SN-8 in which the
mercaptoethylamino group is fixed by introduction of a phenyl ring was
designed. In these compounds with one thiol group, complex oxidative
products are expected to not be produced. Thus, we did not need to
change the thiol group to a disulfide.
SN-5 and SN-6 were synthesized with the same method as SN-214
(Scheme 1). The corresponding aryl dialdehyde was reductively ami-
nated with S-tert-butylcysteamine hydrochloride in the presence of so-
dium cyanoborohydride to give compounds 1 and 3. Reacting 1/3 with
2-nitrophenylsulfenyl chloride followed by selective release of the ni-
trophenylsulfenyl moiety from the amino groups afforded SN-5/SN-6
2

M.O. Radwan, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Scheme 1. Synthetic route of SN-5 and SN-6.
Scheme 2. Synthetic route of SN-9.
via 2/4 (Scheme 1). The overall yields of SN-5 and SN-6 were 3.7% and
13%, respectively. SN-7 was obtained by a straightforward reaction
between 2-(bromomethyl)pyridine hydrobromide and 2-aminoetha-
nethiol hydrochloride in the presence of sodium bicarbonate (94%
yield) as reported previously.19 To obtain SN-8, 2-(pyridin-2-yl)-2,3-
dihydrobenzo[d]thiazole was synthesized20 (87% yield), and then its
thiazoline ring was reductively opened with sodium borohydride21
(70% yield), as previously described. SN-9 was synthesized by reductive
amination of 4-(dimethylamino)picolinaldehyde with S-tritylcystea-
mine hydrochloride in the presence of triacetoxyborohydride (18%
yield), followed by removal of the trityl group (22% yield) (Scheme 2).
Whereas the yields of reductive amination products were generally low,
the final compounds were obtained in good purity. Additionally, we
included the previously reported compounds Biotin-SN-1,17 Farnesyl-
SN-1, and Dodecyl-SN-122 in our study to obtain more thorough
structure-activity relationship information (Fig. 1).
observed. SN-2 and SN-3 are expected to generate the same SN-1 by
reductive agents inside cells, and thus, we deduced that SN-3 is a
mixture of large molecules formed by oxidation of thiol groups. To
correctly evaluate the activity of the reductive type of these com-
pounds, DL-dithiothreitol (DTT) (30 μM) was added to the compounds
first, and the experiment was repeated using the same conditions.
As shown in Fig. 2B, SN-3 showed similar activity (85%) to that of
SN-2 (89%), demonstrating that this assay system is valid. Compared
with SN-2, SN-6 had similar (84%) but slightly weaker activity,
showing the small effect of a dimethylamino group on the pyridine ring.
SN-5 did not show clear inhibitory activity. This demonstrates the im-
portance of the nitrogen atom in the pyridine ring. Notably, 2-ni-
trobenzenethiol (NpSH), which is generated from Bis(2-nitrophenyl)
disulfide and is also formed by reductive cleavage of SN-2, SN-5, and
SN-6, did not show activity. SN-7, SN-8, and SN-9 lacked any detectable
activity, implying the importance of the two side chains of the pyridine
ring. Among Biotin-SN-1, Farnesyl-SN-1, and Dodecyl-SN-1, only
Biotin-SN-1 showed activity (78%), although it was weaker than that of
SN-3. This suggests that a long substituent at position 4 of the pyridine
group is not a major obstacle to activity. The reason that Farnesyl-SN-1
and Dodecyl-SN-1 had no activity may be because these compounds
localize to the membrane or to other proteins with a hydrophobic
pocket. In these experiments, SN-2 showed the best activity to suppress
downstream signaling of TRAF6.
To assess the inhibitory activity of TRAF6, its downstream NF-κB
signaling was examined with a cellular reporter assay.18,23 HeLa S3
cells co-transfected with 3κB-tk-luc (with a triple κB sequence that
binds to NF-κB joined to a downstream firefly luciferase gene) and
control vector pRL-Luc (with a β-actin promoter joined to a down-
stream renilla luciferase gene) were incubated for 1 day, and SN-1 or its
derivative (10 μM) was added. After 1 h of incubation, cells were sti-
mulated with interleukin-1α (IL-1 α) and further incubated for 3 h.
Cells were lysed, and a dual-luciferase reporter assay was conducted.
The value of luminescence after addition of a substrate of firefly luci-
ferase was normalized to that using a substrate of renilla luciferase to
determine NF-κB activity.
The experimental results can be explained by our previous mole-
cular docking study of SN-1 into the first zinc finger of TRAF6 using the
Molecular Operating Environment.18 Both the pyridyl nitrogen and two
sulfurs bind to an amino acid or zinc of TRAF6, showing their im-
portance (binding score −8.5 kcal/mol). Herein, we predicted that SN-
9 interacts in the same shallow pocket. The ionized sulfur interacts with
zinc and Cys155 by ionic and H-bonds, respectively. One more
The result is shown in Fig. 2A. As previously reported, SN-2 showed
a pronounced inhibitory effect (80%) against NF-κB activation.18 SN-6
showed a similar effect (76%), whereas no activity of SN-3 was
3

M.O. Radwan, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Fig. 2. Inhibitory activity against NF-κB activation determined by the reporter assay. HeLa S3 cells were transfected with 3κB-tk-luc and pRL-Luc and incubated for
1 day, followed by treatment with compounds (10 μM) in the absence (A) or presence (B) of DTT (30 μM) and incubated for 1 h. The cells were further treated with IL-
1α (10 ng/mL), and the dual-luciferase reporter assay was performed 3 h post-transfection. The transcription activity from the κB site (in 3κB-tk-luc) was normalized
to the activity from the β-actin promoter (in pRL-Luc). The relative NF-κB activity is shown.
intermolecular bond is formed between the pyridyl moiety and the
His151 residue (binding score −6.2 kcal/mol) (Fig. 3). Visual analysis
and binding score values can provide information about the key role of
the second mercaptoethyl amino chain that conferred superior TRAF6
inhibitory activity on SN-1 by forming a more favorable complex with
the first zinc finger of TRAF6.
contrast, SN-2 did not show cytotoxicity. Both SN-2 and SN-3 are ex-
pected to generate SN-1 due to reductive agents in the cells. The dif-
ference is that only SN-2 will generate NpSH. One plausible mechanism
for the absence of toxicity of SN-2 is that NpSH inhibits the interaction
of SN-2 with various proteins to exert toxicity. However, NpSH itself
shows weak toxicity (24%) (Fig. 4), and thus, the mechanism is not
simple. Importantly, SN-6 showed cytotoxicity unlike SN-2, showing
that the dimethylamino group in the pyridine ring decreases toxicity.
Biotin-SN-1 with a long substituent had similar toxicity to that of SN-3
with a dimethylamino group.
Finally, the cytotoxicity of select compounds was examined. HeLa
S3 cells were incubated with SN-2, SN-3, SN-6, Biotin-SN-1, or bis(2-
nitrophenyl)disulfide with anti-NF-κB activity that were pretreated
with DTT. An MTT [3-(4,5-Dimethylthial-2-yl)-2,5-Diphenyltetrazalium
Bromide] assay was performed after 8 h of incubation, which was a
longer time than that in the experiment to determine NF-κB activity. As
shown in Fig. 4, SN-3 induced death in around half of the cells. In
Taken together, our results suggest that SN-2 has the structure with
the best activity to suppress downstream signaling by binding to
TRAF6. Especially in SN-2, the nitrogen atom of the pyridine ring and
4

M.O. Radwan, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Fig. 3. Binding mode of SN-1 or SN-9 to the first zinc
finger of TRAF6 (PDB ID: 3HCS) predicted by
Molecular Operating Environment. The modeling of
SN-1 was described previously.18 (A) 3D ribbon dia-
gram of SN-9 (left) and SN-1 (right). Hydrogen bonds
between compounds and TRAF6 are shown as black
dashed lines. Amino acids interacting with the com-
pounds are shown in pale blue, and the compounds
are in green. Nitrogen, oxygen, and sulfur are colored
in blue, red, and yellow, respectively. (B) 2D depic-
tion of SN-9 (left) and SN-1 (right). Hydrogen bonds
between compounds and TRAF6 are shown as red
dashed lines. (For interpretation of the references to
colour in this figure legend, the reader is referred to
the web version of this article.)
experiments. This study may become the basis for a new category of
drugs targeting zinc fingers. We have already succeeded in adding
target specificity to SN-1 by changing its structure15 or introducing a
moiety to selectively bind to target proteins22,24. Development of
TRAF6-specific inhibitors based on SN-2 is in progress.
Acknowledgements
We thank Dr. Shigeki Miyamoto (The University of Wisconsin) for
donating the 3κB-tk-luc vector. This work was supported by the Grant
for Joint Research Project of the Institute of Medical Science, the
University of Tokyo.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.bmcl.2019.06.050.
Fig. 4. Cytotoxic activity of compounds. HeLa S3 cells were treated with
compounds (10 μM) in the presence of DTT (30 μM). After 8 h of incubation, the
MTT assay was performed. The relative activity is shown.
References
1. Coleman JE. Zinc proteins: enzymes, storage proteins, transcription factors, and re-
plication proteins. Annu Rev Biochem. 1992;61:897–946.
2. Knaus T, Uhl MK, Monschein S, Moratti S, Gruber K, Macheroux P. Structure and
stability of an unusual zinc-binding protein from Bacteroides thetaiotaomicron. BBA.
2014;1844(12):2298–2305.
its two side chains are required for the activity. The dimethylamino
group of the pyridine ring and the nitrophenylsulfenyl group are im-
portant for decreasing cytotoxicity. We conclude that SN-2 has the
minimum structure for an inhibitor of TRAF6 as shown with the cellular
3. Iannuzzi C, Adrover M, Puglisi R, Yan R, Temussi PA, Pastore A. The role of zinc in
the stability of the marginally stable IscU scaffold protein. Protein Sci.
2014;23(9):1208–1219.
5

M.O. Radwan, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
4. Yamamoto M, Koga R, Fujino H, et al. Zinc-binding site of human immunodeficiency
virus 2 Vpx prevents instability and dysfunction of the protein. J Gen Virol.
2017;98(2):275–283.
involved in the signaling pathway of the replication of human immunodeficiency
virus 1. Bioorg Med Chem. 1997;5(1):205–215.
16. Ejima T, Hirota M, Mizukami T, Otsuka M, Fujita M. An anti-HIV-1 compound that
increases steady-state expression of apoplipoprotein B mRNA-editing enzyme-cata-
lytic polypeptide-like 3G. Int J Mol Med. 2011;28(4):613–616.
17. Radwan MO, Sonoda S, Ejima T, et al. Zinc-mediated binding of a low-molecular-
weight stabilizer of the host anti-viral factor apolipoprotein B mRNA-editing enzyme,
catalytic polypeptide-like 3G. Bioorg Med Chem. 2016;24(18):4398–4405.
18. Koga R, Radwan MO, Ejima T, et al. A dithiol compound binds to the zinc finger
protein TRAF6 and suppresses its ubiquitination. ChemMedChem.
2017;12(23):1935–1941.
5. Koga R, Yamamoto M, Ciftci HI, Otsuka M, Fujita M. Introduction of H2C2-type zinc-
binding residues into HIV-2 Vpr increases its expression level. FEBS Open Bio.
2018;8(1):146–153.
6. Inoue J, Gohda J, Akiyama T. Characteristics and biological functions of TRAF6. Adv
Exp Med Biol. 2007;597:72–79.
7. Walsh MC, Lee J, Choi Y. Tumor necrosis factor receptor- associated factor 6 (TRAF6)
regulation of development, function, and homeostasis of the immune system.
Immunol Rev. 2015;266(1):72–92.
8. Deng L, Wang C, Spencer E, et al. Activation of the IkappaB kinase complex by TRAF6
requires a dimeric ubiquitin-conjugating enzyme complex and a unique polyubiquitin
chain. Cell. 2000;103(2):351–361.
19. De Almeida MV, Chaves JDS, Fontes APS, César ET, Gielen M. Synthesis and char-
acterization of platinum(II) complexes from trifluoromethyl phenylenediamine, pi-
coline and N-benzyl ethylenediamine derivatives. J Braz Chem Soc.
2006;17(7):1266–1273.
9. Sun L, Deng L, Ea CK, Xia ZP, Chen ZJ. The TRAF6 ubiquitin ligase and TAK1 kinase
mediate IKK activation by BCL10 and MALT1 in T lymphocytes. Mol Cell.
2004;14(3):289–301.
20. Carlson LJ, Welby J, Zebrowski KA, et al. Spectroscopic differences between het-
erocyclic benzothiazoline, -thiazole and imine containing ligands and comparison of
the Co and Cu pyridine benzothiazole and imine complexes. Inorg Chim Acta.
2011;365(1):159–166.
10. Yamazaki K, Gohda J, Kanayama A, et al. Two mechanistically and temporally dis-
tinct NF-kappaB activation pathways in IL-1 signaling. Sci Signal. 2009;2(93):ra66.
11. Lamothe B, Campos AD, Webster WK, Gopinathan A, Hur L, Darnay BG. The RING
domain and first zinc finger of TRAF6 coordinate signaling by interleukin-1, lipo-
polysaccharide, and RANKL. J. Biol. Chem. 2008;283(36):24871–24880.
12. Supuran CT, Winum JY, eds. Drug design of zinc-enzyme inhibitors: functional, struc-
tural, and disease applications. Wiley; 2009.
21. Brand U, Vahrenkamp H. Zinc complexes of the N, N, S ligand (mercaptophenyl)
(picolyl)amine. Chem Ber. 1996;129(4):435–440.
22. Tanaka A, Radwan MO, Hamasaki A, et al. A novel inhibitor of farnesyltransferase
with a zinc site recognition moiety and a farnesyl group. Bioorg Med Chem Lett.
2017;27(16):3862–3866.
13. Cushman DW, Ondetti MA. Design of angiotensin converting enzyme inhibitors. Nat
Med. 1999;5(10):1110–1113.
23. Kanemaru Y, Momiki Y, Matsuura S, et al. An artificial copper complex incorporating
a cell-penetrating peptide inhibits nuclear factor-kappaB (NF-kappaB) activation.
Chem Pharm Bull. 2011;59(12):1555–1558.
14. Fujita M, Otsuka M, Sugiura Y. Metal-chelating inhibitors of a zinc finger protein
HIV-EP1. Remarkable potentiation of inhibitory activity by introduction of SH
groups. J Med Chem. 1996;39(2):503–507.
24. Hamasaki A, Naka H, Tamanoi F, Umezawa K, Otsuka M. A novel metal-chelating
inhibitor of protein farnesyltransferase. Bioorg Med Chem Lett.
2003;13(9):1523–1526.
15. Otsuka M, Fujita M, Sugiura Y, et al. Synthetic inhibitors of regulatory proteins
6