I like finding unexpected isosteric relationships. An example that was published a couple of years ago demonstrated that under certain circumstances, pyridine can be used as an isosteric replacement for a carboxylic acid, in contrast to the usual heterocycles that are often used for this purpose..tetrazole, triazole, etc… In this case, Merck was examining a series of diketoacid based compounds ,such as L-731,988, as inibitors of HIV-integrase. One doesn’t have to be a highly experienced medicinal chemist to see the potential pitfalls with this chemotype…there are a number of ways that these compounds might covalently bind to protein increasing the chances for an idiosyncratic reaction. Therefore, an alternative chemotype was needed. In order to find a drug-like chemotype they needed to figure out how to mimic the pharmacophore with “safer” functionality. The first step was to get rid of the carboxylic acid and to do this they took advantage of proposed mechanism of actino of the diketoacids. When fully deprotonated, the diketoacid is seen to resemble the pentameric intermediate/transition state of a phosphate ester bond cleavage, suggesting that this chemotype can bind to the Mg+2 metals located in the active site. According to this hypothesis the carboxylic acid functions as a metal binding element. This being the case they found that replacing it with pyridine retained activity. In doing so, they deomstrate that a basic heterocycle can be used to replace a carboxylic acid. Where they went from here can be found in a J. Med. Chem. paper.