Hit-to-Lead optimisation is a crucial step in Drug Discovery. It implies the wise modification of hit molecules by following specific pharmacologic and pharmacokinetic parameters. Many strategies can be employed to tackle this challenge, one of them is bioisosterism. Bioisosters are moieties or atoms that show the same physicochemical properties and biological activity. Thus, medicinal chemists can rely on a large chemical toolbox, for example, by changing an amide bond to an oxazole or shielding a carboxylic acid with a tetrazole. This all depends how we want to drive the series (in terms of physicochemical properties) through the bottleneck of the Drug Development. Bioisosterism is also widely used for the IP space expansion of chemical libraries.
In this light, I would like to discuss the “big brother” of carbon, the silicon. Within the third row of the periodic table, silicon is located below the carbon; they share the same valency of 4 and commonly forms tetrahedral molecules, the most common silicon linkage being Si-C and Si-O. The replacement of carbon by silicon within bioactive compounds could therefore yield new compounds with different properties for lead optimization.1 Small chemical differences exist between silicon and carbon. Indeed, it is known that the C-Si bond is 20% larger than the C-C bond – this observation has consequences on the shape and conformation of the molecule, which in turn leads to different interactions with the biological system. Silicon compounds are also more lipophilic than their carbon congeners. Therefore, switching from carbon to silicon could improve cell penetration, which is very important for compounds targeting the central nervous system for example. Nevertheless it also creates solubility and metabolic clearance issues that could be mitigated, depending where we want to put the cursor in terms of DMPK. A “hidden” feature of silicon is that it can form hexacoordinated compounds in comparison with carbon: that has great significance in Medicinal Chemistry since many potent transition state mimics containing silanediols have been developed. Finally, silicon is more electropositive than carbon, which leads to a difference in bond polarity and ultimately to a different biological outcome, one good example is the ammonium/silicon exchange found in Zifrosilone (acetylcholinesterase inhibitor).
A lot of work have been produced recently towards the pharmacological evaluation of new silicon-containing molecules (Figure 1), however none of these progresses has yet yielded a marketed drug. As said recently by the blogger Derek Low,2 silicon stays in the shadows, despite the huge potential offered by this element in balancing physicochemical properties with DMPK and lowering compound attrition during the lead optimization phase.
I believe that a new era for silicon in Drug Discovery will come soon; we cannot neglect this element any longer.
Figure 1. Examples of some bioactive silicon-containing molecules with enhanced pharmacology and DMPK
Blog written by Mohamed Benchekroun
(1) Ramesh, R.; Reddy, D. S. Quest for Novel Chemical Entities through Incorporation of Silicon in Drug Scaffolds. J. Med. Chem. 2017.
(2) Lowe, D. Silicon Stays in the Shadows http://blogs.sciencemag.org/pipeline/archives/2017/11/07/silicon-stays-in-the-shadows (accessed Dec 7, 2017).