Displaying enzymatic IC 50 values of 19 nM and less than 15 nM against GLP and G9a,
respectively [10], the inhibitor 4 is more or less equipotent against G9a/GLP. Kinetic
studies in the presence of histone and SAM, as well as a crystal structure analysis,
confirmed that compound 4 is a substrate-competitive inhibitor [10]. Compound
4 exhibited high cellular potency and low cell toxicity in several cancer and normal
cell lines. For example, it lowered global H3K9me2 levels in human breast cancer
MDA-MB-231 cells, being significantly more potent (IC 50 : 81 Æ 9 nM) than compound 1 (IC 50 : 500 Æ 43 nM) [10]. As mentioned before, compound 4 was less
cytotoxic (EC 50 : 11,000 Æ 710 nM) than compound 1 (EC 50 : 2,700 Æ 76 nM) in
MDA-MB-231 cells, thus offering a better therapeutic window. Additionally, compound 4 has been shown to induce leukemic stem cell differentiation, suppressing the
proliferation of primary human AML cells [11]. The potential clinical benefit of such a
pharmacological inhibition of G9a in AML treatment was confirmed in
G9a-conditional knockout AML cells [11]. The cyclohexyl group present in the
2-position of compound 4 is known to be prone to metabolic modifications, so that a
more suitable and stable candidate for in vivo studies was needed [10]. The same
research group further investigated on the 2-position by using more stable substituents
such as nitrogen-containing saturated heterocycles, leading to excellent biochemical
properties together with IC 50 s against G9 lower than 2.5 nM [12]. UNC0642 (5,
Fig. 1), possessing a 4,4-difluoropiperidine substitution at C-2, succeeded compound
4, being a suitable candidate for in vivo studies [12]. Compound 5 retained high in vitro
potency for G9a and GLP (IC 50 < 2.5 nM) and was >20,000-fold selective for
G9a/GLP over other methyltransferases (e.g., SETD7, SETD8, SETDB1, PRMTs,
SUVs, DOT1L, and DNMT1) and >300-fold selective over other non-epigenetic
targets such as kinases. This compound reduced the H3K9me2 mark, without
exhibiting a strong cytotoxicity, in both healthy and cancer cell lines. Compound
5 might not only be useful in cancer therapy but also in Prader-Willi syndrome (PWS),
where it was proved to reactivate genes from the maternal allele not only in a cellular
context but also in a mouse model, leading to a significantly increased growth and
lifespan of PWS mouse pups [13]. Recently, MS012 (6, Fig. 1), a potent and selective
(over PKMTs, PRMTs, DNMTs, and RNMTs) GLP inhibitor has been described
[14]. This inhibitor possesses still a quinazoline scaffold and exhibits a 140-fold
selectivity for GLP (IC 50 : 7 Æ 2 nM), over G9a. It should be mentioned that the
X-ray structures revealed that this substrate-competitive inhibitor has an almost identical binding mode for GLP and G9a, underlining the challenge for a medicinal chemist
to design potent and highly selective inhibitors for such homologous enzymes
[14]. A-366 (7, Fig. 1), the first potent and selective non-quinazoline G9a inhibitor,
has been described by Abbvie (IL, USA) featuring a novel spiro(cyclobutane1,3
0 -indol)-2
0 -amine core preserving the right-hand side of the UNC G9a inhibitors.
This compound potently inhibited G9a (IC 50 : 3.3 nM) and GLP (IC 50 : 38 nM) with a
substrate-competitive and non-SAM-competitive mechanism of action (MOA)
[15]. The crystal structure of compound 7 in complex with G9a revealed a comparable
binding mode to compound 4 [15]. Like compound 4, compound 7 at 3 μM reduced by
50% H3K9me2 levels in prostate cancer cells (PC-3), although no effects on cellular
proliferation have been observed on a panel of 38 cancer cell lines even at 10 μM.
128
G. Stazi et al.
respectively [10], the inhibitor 4 is more or less equipotent against G9a/GLP. Kinetic
studies in the presence of histone and SAM, as well as a crystal structure analysis,
confirmed that compound 4 is a substrate-competitive inhibitor [10]. Compound
4 exhibited high cellular potency and low cell toxicity in several cancer and normal
cell lines. For example, it lowered global H3K9me2 levels in human breast cancer
MDA-MB-231 cells, being significantly more potent (IC 50 : 81 Æ 9 nM) than compound 1 (IC 50 : 500 Æ 43 nM) [10]. As mentioned before, compound 4 was less
cytotoxic (EC 50 : 11,000 Æ 710 nM) than compound 1 (EC 50 : 2,700 Æ 76 nM) in
MDA-MB-231 cells, thus offering a better therapeutic window. Additionally, compound 4 has been shown to induce leukemic stem cell differentiation, suppressing the
proliferation of primary human AML cells [11]. The potential clinical benefit of such a
pharmacological inhibition of G9a in AML treatment was confirmed in
G9a-conditional knockout AML cells [11]. The cyclohexyl group present in the
2-position of compound 4 is known to be prone to metabolic modifications, so that a
more suitable and stable candidate for in vivo studies was needed [10]. The same
research group further investigated on the 2-position by using more stable substituents
such as nitrogen-containing saturated heterocycles, leading to excellent biochemical
properties together with IC 50 s against G9 lower than 2.5 nM [12]. UNC0642 (5,
Fig. 1), possessing a 4,4-difluoropiperidine substitution at C-2, succeeded compound
4, being a suitable candidate for in vivo studies [12]. Compound 5 retained high in vitro
potency for G9a and GLP (IC 50 < 2.5 nM) and was >20,000-fold selective for
G9a/GLP over other methyltransferases (e.g., SETD7, SETD8, SETDB1, PRMTs,
SUVs, DOT1L, and DNMT1) and >300-fold selective over other non-epigenetic
targets such as kinases. This compound reduced the H3K9me2 mark, without
exhibiting a strong cytotoxicity, in both healthy and cancer cell lines. Compound
5 might not only be useful in cancer therapy but also in Prader-Willi syndrome (PWS),
where it was proved to reactivate genes from the maternal allele not only in a cellular
context but also in a mouse model, leading to a significantly increased growth and
lifespan of PWS mouse pups [13]. Recently, MS012 (6, Fig. 1), a potent and selective
(over PKMTs, PRMTs, DNMTs, and RNMTs) GLP inhibitor has been described
[14]. This inhibitor possesses still a quinazoline scaffold and exhibits a 140-fold
selectivity for GLP (IC 50 : 7 Æ 2 nM), over G9a. It should be mentioned that the
X-ray structures revealed that this substrate-competitive inhibitor has an almost identical binding mode for GLP and G9a, underlining the challenge for a medicinal chemist
to design potent and highly selective inhibitors for such homologous enzymes
[14]. A-366 (7, Fig. 1), the first potent and selective non-quinazoline G9a inhibitor,
has been described by Abbvie (IL, USA) featuring a novel spiro(cyclobutane1,3
0 -indol)-2
0 -amine core preserving the right-hand side of the UNC G9a inhibitors.
This compound potently inhibited G9a (IC 50 : 3.3 nM) and GLP (IC 50 : 38 nM) with a
substrate-competitive and non-SAM-competitive mechanism of action (MOA)
[15]. The crystal structure of compound 7 in complex with G9a revealed a comparable
binding mode to compound 4 [15]. Like compound 4, compound 7 at 3 μM reduced by
50% H3K9me2 levels in prostate cancer cells (PC-3), although no effects on cellular
proliferation have been observed on a panel of 38 cancer cell lines even at 10 μM.
128
G. Stazi et al.