A lot has been said about azide/alkyne cycloaddition reaction in recent years. This process (click chemistry) has been applied in a wide range of research fields and the spectrum of use keeps expanding. One particularly widely quoted tenet of this chemistry states that azides and alkynes are orthogonal to polar functionalities, but react with high selectivity with each other under copper catalysis. This is especially useful in the so-called bioorthogonal applications. While this principle is mostly true, it is certainly not universal, which is why it is instructive to see cases that disprove the basic assumption of alkynes’ “innocence”. Today I want to mention just one example. I have always been intrigued by the mechanism of covalent inhibition of Erb kinases that was elucidated by David Uehling and coworkers form GSK some years ago. This paper was published in PNAS and you can check it out below. Granted, the core of this particular line of inhibitors might be described as a fairly special alkyne unit that is perhaps more prone to nucleophilic addition than some others. Nonetheless, one should keep in mind that alkynes are not always orthogonal to nucleophilic species in biological systems.
http://www.pnas.org/content/105/8/2773.abstract
amphoteros 
If you want to avoid copper-catalyzed click, i.e. in bio-compatible ligation with nucleic acids and proteins in aqueous buffer at room temperature, the chances are you are using strained cyclooctyne- or trans-cyclooctene modified ligand. Which means you cannot use solid phase synthesis all the way (because cyclooctyne does not like TFA), and the final product is likely to go bad on air in one day – unless stored at -80C. Just putting a cyclooctyne like DAIBO or BCN on unprotected peptide sidechain and purifying the resulting mix is a major challenge. There goes the orthogonality.
Great point, thanks.