Synthesis of novel derivatives of 5-hydroxymethylcytosine and 5-formylcytosine as tools for epigenetics

Summary In this work we present for the first time the synthesis of novel 5-hydroxymethylcytosine (5hmC) and 5-formylcytosine (5fC) derivatives that can be used as tools in the emerging field of epigenetics for deciphering chemical biology of TET-mediated processes.

(3) Among other putative demethylation mechanisms is the direct dehydroxymethylation of 5hmC to cytosine by action of DNA methyltransferases (DNMT). This enzymatic process was observed in vitro, whether it also works in vivo is yet to be elucidated [15,22]. (4) Lastly, decarboxylation of the 5caC by an unknown decarboxylase excluding action of BER should also be considered [15,23]. This variety of demethylation pathways might indicate that different tissues utilize different demethylation pathways [1,24].
While DNA methylation is usually associated with gene repression [8,25], active demethylation seems to allow cells to unblock silenced genes aiming at epigenetic reprogramming of their genetic material [26]. Current accepted models propose that 5hmC could be involved in epigenetic modulation of gene activity. In fact, 5hmC was discovered also in embryonic stem cells and seems to play a decisive role in their self-renewal process [27]. Interestingly, the levels of 5hmC in several cancer types are strongly reduced relative to the corresponding normal tissue around the tumor [28]. To gain deeper insights into the chemical biology of DNA demethylation pathways further exploration of the TET-mediated processes is necessary. Analogues of 5hmC with substituents preventing formation of 5fC and 5caC species could serve as useful tools for ongoing investigations in this emerging field.

Results and Discussion
Herein, we describe the synthesis of compounds with the general formula I which represents modified cytidine analogues bearing a secondary alcohol at position C-5 of cytosine. Additionally, a synthesis of 3,6-dihydrodeoxycytidine derivatives of general formula II is presented (Figure 1). We chose the known aldehyde 1 [29] (prepared from commercially available 2'-deoxycytidine) as a starting material for the envisioned transformations (Scheme 2). To the best of our knowledge, the addition of organometallic compounds (organolithium and organomagnesium, etc.) to aldehyde 1 is not described in the literature. Compound 1 was readily converted to 5hmC analogues 2a-e by treatment with various Grignard reagents (methylmagnesium bromide, THF, 0 °C → room temperature, or vinylmagnesium bromide, THF, 0 °C → room temperature) and organolithium reagents (lithium (trimethylsilyl)acetylide, THF, −40 °C → −20 °C or lithium phenylacetylide, THF, −78 °C → −50 °C) (Scheme 2). These alcohols were obtained as a mixture of diastereomers in yields ranging from 43% to 96% (Table 1). Compound 2b was isolated in moderate yield of 43% due to the cleavage of the TMS-group during the reaction resulting in formation of derivative 2e with a yield of 26%. The obtained derivatives 2a-d were further  treated with Olah's reagent and pyridine in EtOAc at room temperature or HF·triethylamine complex [30] in DCM at 0 °C to afford the deprotected 2'-deoxycytidine analogues 3a-d as mixtures of diastereomers in yields of 60-75%.

Conclusion
In summary, the reaction of 5fC derivatives 1, 4, and 5 with organometallic reagents (RMgBr, RLi) was investigated and enabled the synthesis of novel derivatives of 5-hydroxymethylcytosine and 5-formylcytosine: whereas aldehydes 1 and 4 afforded cytosine derivatives 2a-e, 6a-c and 7a-d, the reaction of derivative 5 yielded 3,6-dihydrodeoxycytidine derivatives 8a-d which subsequently after removal of the TCBoc group afforded derivatives 9a-d. These new nucleobase modified 2'-deoxycytidine analogues can be used in the synthesis [29,33,34] of modified DNA oligomers for further studies of the TET-mediated processes which are of great importance in the emerging field of epigenetics. In addition they could find application as novel antivirals and/or as antimetabolites [35,36]. The majority of the obtained compounds contain functionalized side chains thus allowing further manipulations.

Supporting Information
Supporting Information File 1 Experimental details and analytical data of all synthesized compounds.