In 2011, Epizyme Inc. described the first selective DOT1L inhibitor EPZ004777
(36, Fig. 4), based on the cofactor SAM and the crystal structure of the enzyme active
site. This compound possesses a high picomolar potency in an enzyme-based assay
(IC 50 : 400 Æ 100 pM) [82]. It shows a remarkable selectivity (1,000-fold selective)
over other SAM-dependent methyltransferases despite the structural similarity of the
cofactor binding pocket. As expected, binding studies and crystal structures of this
inhibitor within DOT1L confirmed it as a SAM competitor [83]. Reduced levels of
H3K79me2 were found in several leukemia cell lines treated with compound 36,
without affecting other histone methylation marks, thus underlining the specificity of
this compound [82]. Furthermore, compound 36 led to a reduced expression of the
MLL-rearranged leukemia hallmarks HOXA9 and MEIS1 [82], being able to selectively kill at low micromolar potencies MLL-rearranged leukemia cells, while having
little effect on non-MLL-translocated cells, and prolonging survival in a mouse model
of MLL-rearranged leukemia [84]. These effects can be taken as a proof of concept
for the pharmacological inhibition of DOT1L in specific cancer therapy [82]. However, despite these first promising cell-based results, further clinical development was
not pursued as the pharmacokinetic properties of this compound were not ideal.
In 2013, Epizyme Inc. published the second generation of an improved derivative
of compound 36, known as EPZ-5676 or pinometostat (37, Fig. 4) [85]. In this
compound the ribose moiety was replaced with a cyclobutyl ring to enhance its
pharmacokinetic properties as well as its inhibition activity against DOT1L
(K i < 0.08 nM; EPZ004777 K i : 0.3 nM) [85]. Compound 37 retained the same
binding mode as its parent compound. Interestingly, the selectivity over other protein
methyltransferases has been further increased up to 37,000-fold [85]. The new
analogue displayed a nanomolar antiproliferative activity against most of the other
MLL-rearranged cell lines such as MV4–11 (MLL-AF4), MOLM-13 (MLL-AF9),
and THP1 (MLL-AF9), with little effect on leukemia cells lacking the MLL translocation [82, 85]. Despite improved pharmacokinetic profile [86], compound 37 is still
poorly orally bioavailable [87]. To summarize, compound 37 is a major breakthrough
in the protein methyltransferase (PMT) inhibitor field, which entered as the first PMT
inhibitor in the clinical arena: phase I clinical trials for this inhibitor were recently
completed for the treatment of patients with MLL-r, a genetically defined type of
acute leukemia [56]. However, the outcome of this study has not yet been published.
Furthermore, there are also several other DOT1L inhibitors possessing a
deazadenosine core reported in the literature. Using the crystal structure of the
DOT1L-36 complex, Yu et al. developed a chemical probe called SGC0946 (38,
Fig. 4), just showing an additional bromo atom at the 7-position of the deazadenosine
ring of 36, with enhanced in vitro (IC 50 : 0.3 nM) and in vivo potency. This compound
was also able to reduce H3K79 methylation levels (IC 50 : 8.8 Æ 1.6 nM) tenfold more
potently than the lead 36 (IC 50 : 84 Æ 20 nM) in MCF10A cells while maintaining a
good selectivity profile [83]. Yao et al. described compound 39 (Fig. 4) as a selective,
covalent, and potent DOT1L inhibitor (IC 50 : 38 nM), possessing greater than 29-fold
selectivity for DOT1L over other methyltransferases such as CARM1, PRMT1, G9a,
and SUV39H1 [88]. The authors proposed, additionally to the hydrogen bond
interactions within the SET domain, an intramolecular cyclization of compound 39
136
G. Stazi et al.
(36, Fig. 4), based on the cofactor SAM and the crystal structure of the enzyme active
site. This compound possesses a high picomolar potency in an enzyme-based assay
(IC 50 : 400 Æ 100 pM) [82]. It shows a remarkable selectivity (1,000-fold selective)
over other SAM-dependent methyltransferases despite the structural similarity of the
cofactor binding pocket. As expected, binding studies and crystal structures of this
inhibitor within DOT1L confirmed it as a SAM competitor [83]. Reduced levels of
H3K79me2 were found in several leukemia cell lines treated with compound 36,
without affecting other histone methylation marks, thus underlining the specificity of
this compound [82]. Furthermore, compound 36 led to a reduced expression of the
MLL-rearranged leukemia hallmarks HOXA9 and MEIS1 [82], being able to selectively kill at low micromolar potencies MLL-rearranged leukemia cells, while having
little effect on non-MLL-translocated cells, and prolonging survival in a mouse model
of MLL-rearranged leukemia [84]. These effects can be taken as a proof of concept
for the pharmacological inhibition of DOT1L in specific cancer therapy [82]. However, despite these first promising cell-based results, further clinical development was
not pursued as the pharmacokinetic properties of this compound were not ideal.
In 2013, Epizyme Inc. published the second generation of an improved derivative
of compound 36, known as EPZ-5676 or pinometostat (37, Fig. 4) [85]. In this
compound the ribose moiety was replaced with a cyclobutyl ring to enhance its
pharmacokinetic properties as well as its inhibition activity against DOT1L
(K i < 0.08 nM; EPZ004777 K i : 0.3 nM) [85]. Compound 37 retained the same
binding mode as its parent compound. Interestingly, the selectivity over other protein
methyltransferases has been further increased up to 37,000-fold [85]. The new
analogue displayed a nanomolar antiproliferative activity against most of the other
MLL-rearranged cell lines such as MV4–11 (MLL-AF4), MOLM-13 (MLL-AF9),
and THP1 (MLL-AF9), with little effect on leukemia cells lacking the MLL translocation [82, 85]. Despite improved pharmacokinetic profile [86], compound 37 is still
poorly orally bioavailable [87]. To summarize, compound 37 is a major breakthrough
in the protein methyltransferase (PMT) inhibitor field, which entered as the first PMT
inhibitor in the clinical arena: phase I clinical trials for this inhibitor were recently
completed for the treatment of patients with MLL-r, a genetically defined type of
acute leukemia [56]. However, the outcome of this study has not yet been published.
Furthermore, there are also several other DOT1L inhibitors possessing a
deazadenosine core reported in the literature. Using the crystal structure of the
DOT1L-36 complex, Yu et al. developed a chemical probe called SGC0946 (38,
Fig. 4), just showing an additional bromo atom at the 7-position of the deazadenosine
ring of 36, with enhanced in vitro (IC 50 : 0.3 nM) and in vivo potency. This compound
was also able to reduce H3K79 methylation levels (IC 50 : 8.8 Æ 1.6 nM) tenfold more
potently than the lead 36 (IC 50 : 84 Æ 20 nM) in MCF10A cells while maintaining a
good selectivity profile [83]. Yao et al. described compound 39 (Fig. 4) as a selective,
covalent, and potent DOT1L inhibitor (IC 50 : 38 nM), possessing greater than 29-fold
selectivity for DOT1L over other methyltransferases such as CARM1, PRMT1, G9a,
and SUV39H1 [88]. The authors proposed, additionally to the hydrogen bond
interactions within the SET domain, an intramolecular cyclization of compound 39
136
G. Stazi et al.