P. Jaikhan, et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Fig. 3. Estimation of Gibbs free energies (ΔG) in complex formation between
ortho-amino anilide and zinc/iron ion by a theoretical quantum calculation
study with Gaussian 09.
Fig. 1. Structures of representative KDM inhibitors.
are required as cell-active KDM inhibitors. Although CPI-455 (6,
Fig. 1),17 a novel KDM inhibitor, exhibited relatively high activity in a
cellular assay, the number of cell-active αKG mimic inhibitors remains
limited. In this study, we attempted to identify a novel αKG mimic
scaffold for Fe(II)/αKG-dependent KDM inhibitors.
inhibitors. Ortho-substituted anilides 9 and 10 were designed: acet-
amide and formamide substitutions represent two different species of
the anilide core, and a hydroxy group was also used as an ortho-sub-
stituted group as well as an amino group (Fig. 2B).
The designed ortho-anilides were synthesized from corresponding
benzoic acids, as outlined in Scheme 1. Esterification of compounds 11,
13, and 15 afforded methyl esters 12, 14, and 16, respectively. Com-
pounds 12, 14, and 16 were treated with acetyl chloride to yield
compounds 19, 21, and 22, respectively. Compound 19 was converted
into amine 20 by hydrogenation in the presence of Pd/C. Finally, de-
protection of compounds 20–22 under the basic condition gave desired
compounds 9a, 9b, and 10a, respectively. On the other hand, com-
pounds 9c and 10b were prepared by the direct formamidation of
compounds 17 and 18, respectively. The final compounds were purified
by chromatography and their purities (> 95%) were confirmed by
HPLC analysis.
We focused on an ortho-substituted anilide structure. Various ortho-
amino anilides have been reported as histone deacetylase (HDAC) in-
hibitors, such as MS-275 (7) and T247 (8) (Fig. 2A).32 The ortho-amino
anilide moiety of these inhibitors can chelate a zinc ion in the HDAC
catalytic site, inhibiting HDACs strongly. Based on the metal chelating
mechanism of the ortho-amino anilide group of HDAC inhibitors, we
thought that ortho-substituted anilide would be used as a novel phar-
macophore of KDM inhibitors. In other words, we expected that the
ortho-substituted anilide scaffold would be useful as an iron chelator,
and that ortho-substituted anilides bearing a carboxylate group would
work as an αKG competitor (Fig. 2B). Before we experimentally verified
our hypotheses, we explored the possibility that an ortho-substituted
anilide would work as an iron chelator by performing a theoretical
quantum calculation study with Gaussian 09.33 We estimated the
complex formation energies between an ortho-substituted anilide and a
zinc ion or an iron ion, and compared them (Fig. 3). To simplify the
calculations, we used water molecules as the metal ligand (Fig. 3).
lent and hexavalent, respectively,34,35,36 we optimized the structures of
a pentavalent zinc complex and a hexavalent iron complex. The geo-
metry optimization and the vibration frequency of the ortho-substituted
anilide and its complexes with the metals were calculated using a
density functional model (B3LYP) with the 6–31+G* basis set. As a
result, the Gibbs free energies (ΔG) for the formation of amino anilide-
Zn2+, amino anilide-Fe2+, and hydroxy anilide-Fe2+ complexes were
estimated to be −31.94 kcal/mol, −30.12 kcal/mol, and −26.46 kcal/
mol, respectively, and there was little difference between them. These
calculation results strongly suggested that the ortho-substituted anilides
should coordinate not only with Zn2+ but also with Fe2+. Therefore, we
attempted to evaluate ortho-substituted anilides as novel KDM
The activities of the ortho-substituted anilides were initially
screened against KDM5A (also known as JARID1A) enzyme in the
AlphaScreen™ assay. NCDM-81a (5), a KDM5A inhibitor, was used as
positive control. Although amino-substituted acetanilide 9a did not
inhibit KDM5A at 50 µM, hydroxy-substituted acetanilide 9b showed
weak inhibitory activity against KDM5A at the same concentration
(Fig. 4), suggesting that the hydroxy group is preferred to the amino
group in KDM5A inhibition. Then, we tested hydroxy-substituted
acetanilides 10 bearing a 4-carboxylate group. The KDM5A inhibitory
activity of 10b was significantly higher than that of 10a, which in-
dicates that the formanilide moiety is better than the acetanilide
moiety. Accordingly, we tested hydroxy-substituted formanilide 9c
bearing a 3-carboxylate group, and found that 9c exhibited pronounced
KDM5A inhibitory activity. These results suggest that the 3-carbox-
ylates, if anything, are preferred to the 4-carboxylates in the KDM5A
inhibition by the ortho-hydroxy anilides. A superimposition of a stable
conformation of an ortho-hydroxy anilide coordinating Fe2+ ion with
the conformation of αKG bound to KDM5A indicates that the 3-position
Fig. 2. (A) Structures of representative ortho-amino anilide HDAC inhibitors, MS-275 (7) and T247 (8). (B) Design of ortho-substituted anilides for KDM inhibitors.
2