A new synthesis for cyclic diguanylic acid and its analogues

Cyclic bis(3’-5’)diguanylic acid (c-di-GMP) has been the focus of many research
endeavors for biologists in the last decade. Indeed, this cyclic dinucleotide has been
identified as a novel secondary messenger recently.[8-11, 49] This new discovery caused
increasing interest in the regulation system which involves c-di-GMP. This insight, recently
led to widespread findings about c-di-GMP in other bacteria. The cyclic bis(3’-5’)-nucleotide
has been shown to regulate the transition from motility to sessility in bacteria including
Caulobacter crescentus[15], Escherichia coli and the pathogenic bacteria Pseudomonas
aeroginosa and Salmonella typhimurium[7].
This cyclic dinucleotide also showed an influence on community behavior like biofilm
formation in pathogenic bacteria including Pseudomonas fluorescens[16], Yersinia pestis[17]
and Vibrio cholerae[18]. It is also involved in the inhibition of Staphylococcus aureus cell–cell
interactions and biofilm formation, as well as in the reduction of the virulence of the biofilmforming
strains of the same bacterium in a mouse model of mastitis infection.[19]
These findings suggest that cyclic diguanylic acid might be useful in preventing biofilm
formation on clinically relevant surfaces such as medical devices and potentially, in the
control and treatment of human and animal infection.[17] The biological activity might be even
wider since reports have pointed out that this compound may have anticancer activity.[18]
Thus, c-di-GMP represents an excellent platform for drug design in medicinal chemistry
and especially in the field of antibiotics where compounds with new modes of action are
required. However, the mechanisms of c-di-GMP dependent signalling remain unknown,
mainly because little data is available on c-di-GMP.[8,10] In order to study the biochemistry of
this cyclic dinucleotide more in detail we have started this project dedicated to the synthesis
of c-di-GMP and its analogues.
We intended to develop a synthetic pathway which could afford an efficient, reliable,
flexible and scalable route to synthesize c-di-GMP. At the beginning of this work, the only
reported synthetic route for c-di-GMP was the van Boom et al.[22-23] method starting from
guanosine and using the phosphotriester methodology. This method was the starting point of
our own synthetic investigations, even so the published synthesis needed tedious purification
steps and its length rendered it only moderately suitable for eventual scale-up purposes.
In the course of this work, two more synthetic pathways were reported by Hayakawa et
al.[29] and Jones et al.[30] claiming better yields, easier realization and shorter reaction
sequences. We then decided to apply some of their improvements, by modifying the
guanosine building block to make it less polar but still use the phosphotriester methodology
towards an easier assembly of c-di-GMP. However, no previously described method afforded
large quantities of c-di-GMP.
After having explored the different existing synthetic routes, it quickly became obvious
that we would have to design a new method to obtain this compound in sizeable amounts to
satisfy the demands for the biological investigations. We have decided to adopt a brand new
approach in which we start from ribose building blocks and synthesize a sugar-phosphate
backbone, and to introduce the base at a late stage (Scheme A). Through this route we
anticipate to completely solve the difficulties, generally caused by the 2’-OH protection, by
using the 1’,2’-acetal protecting group.
In order to show the flexibility of our new synthetic route, the synthesis of basemodified
analogues of c-di-GMP was undertaken. The intend was to show that the synthesis
is not specific for purine bases but can be applied to pyrimidine bases as well as non natural
nucleobases, such as xanthine or theophylline for example. Finally, we applied the same
strategy to the synthesis of internucleotide linkage modified analogues.giese