studies are revealing the role of DNA methylation in obesity, allergy, autoimmune
diseases, addiction and inflammation [42, 43].
Nevertheless, despite the established roles of DNA methylation and DNMT1 and
DNMT3s in diseases [42, 44], which DNMT isoform is best to therapeutically target
and in which pathology remains to be determined. Chemical tools can thus be useful
to address these important questions. Another unanswered question is in regards to
Dnmt2. Mammalian Dnmt2 was discovered by its homology with the most conserved C5-DNA methyltransferases [45], but it was shown not to methylate DNA
but rather RNA cytosines, the tRNA
Asp [46, 47], and was subsequently renamed
TRDMT1. It contains only the catalytic domain and interestingly is present in
species that do not have DNMT1 or DNMT3 (such as Drosophila and
Schizosaccharomyces pombe) [25]. Today it is still debated whether in certain
species, Dnmt2 is able to methylate DNA or if its action is limited to RNA
methylation [48, 49].
Thus, while DNMT enzymes are well-studied, many questions remain to be
answered. Chemical tools that specifically inhibit the DNMTs can contribute
towards answering these biological questions.
3 Inhibitors of DNA Methylation
Two families of inhibitors have been identified for DNMTs: the nucleoside analogues, of which 5-azacytidine (5azaC (5)) and 5-aza-deoxycytidine (5azadC(6)) are
approved anti-leukaemia drugs (known as azacitidine and decitabine), and the nonnucleoside analogues, which are composed of very different scaffolds (Fig. 3).
3.1 Cytosine Analogues
5azaC (5) was synthetised as an antimetabolite and was described for its anti-leukaemia
properties in 1964–1965 by Sorm and Vesely [50, 51]. However, it was its impact on
DNA methylation and on the reprogramming of cells by Jones and Taylor [52, 53] that
allowed the understanding of its mechanism of action [54–56]. 5azaC and 5azadC
incorporate into DNA instead of dC. Once the DNMT is bound to the position 6 of the
5azadC, the β-elimination and restauration of the 5–6 double bond cannot occur and the
DNMT is irreversibly trapped on the DNA (suicide complex), inducing its degradation
by the proteasome [57]. The trapping of the DNMT1 was elegantly visualised by using
fluorescent DNMT1 fusions [56].
5azaC (5) was approved by the FDA and then the EMA, together with the deoxy
analogue 5azadC (6) for the treatment of acute myeloid leukaemia (AML),
myelodysplastic syndrome (MDS) and chronic myelomonocytic leukaemia (CMML).
Both 5azaC and 5azadC are chemically unstable (storage and handling is sensitive) and have a very short half-life in patients [58]. Several efforts were made to
262
R. Belle et al.
diseases, addiction and inflammation [42, 43].
Nevertheless, despite the established roles of DNA methylation and DNMT1 and
DNMT3s in diseases [42, 44], which DNMT isoform is best to therapeutically target
and in which pathology remains to be determined. Chemical tools can thus be useful
to address these important questions. Another unanswered question is in regards to
Dnmt2. Mammalian Dnmt2 was discovered by its homology with the most conserved C5-DNA methyltransferases [45], but it was shown not to methylate DNA
but rather RNA cytosines, the tRNA
Asp [46, 47], and was subsequently renamed
TRDMT1. It contains only the catalytic domain and interestingly is present in
species that do not have DNMT1 or DNMT3 (such as Drosophila and
Schizosaccharomyces pombe) [25]. Today it is still debated whether in certain
species, Dnmt2 is able to methylate DNA or if its action is limited to RNA
methylation [48, 49].
Thus, while DNMT enzymes are well-studied, many questions remain to be
answered. Chemical tools that specifically inhibit the DNMTs can contribute
towards answering these biological questions.
3 Inhibitors of DNA Methylation
Two families of inhibitors have been identified for DNMTs: the nucleoside analogues, of which 5-azacytidine (5azaC (5)) and 5-aza-deoxycytidine (5azadC(6)) are
approved anti-leukaemia drugs (known as azacitidine and decitabine), and the nonnucleoside analogues, which are composed of very different scaffolds (Fig. 3).
3.1 Cytosine Analogues
5azaC (5) was synthetised as an antimetabolite and was described for its anti-leukaemia
properties in 1964–1965 by Sorm and Vesely [50, 51]. However, it was its impact on
DNA methylation and on the reprogramming of cells by Jones and Taylor [52, 53] that
allowed the understanding of its mechanism of action [54–56]. 5azaC and 5azadC
incorporate into DNA instead of dC. Once the DNMT is bound to the position 6 of the
5azadC, the β-elimination and restauration of the 5–6 double bond cannot occur and the
DNMT is irreversibly trapped on the DNA (suicide complex), inducing its degradation
by the proteasome [57]. The trapping of the DNMT1 was elegantly visualised by using
fluorescent DNMT1 fusions [56].
5azaC (5) was approved by the FDA and then the EMA, together with the deoxy
analogue 5azadC (6) for the treatment of acute myeloid leukaemia (AML),
myelodysplastic syndrome (MDS) and chronic myelomonocytic leukaemia (CMML).
Both 5azaC and 5azadC are chemically unstable (storage and handling is sensitive) and have a very short half-life in patients [58]. Several efforts were made to
262
R. Belle et al.