At the heart of the method is the installation of a maleimide onto the nanoparticle surface. Since the ligands that are used to form the protective self-assembled monolayer on the gold nanoparticle possess thiols to allow bonding to the metal, this requires an indirect method. Dan approached this by incorporating an azide at the end of one of the ligands. After the ligand shell is formed on the nanoparticle, the azide can easily be converted to a maleimide by reaction with a strained cyclooctyne (copper-free click chemistry, which is useful for biomolecules, since Cu+ is quite reactive). The maleimide can then be reacted with a thiol on a protein (Dan used fibroblast growth factor-2) or indeed any other molecule.
Highlighting the versatility of the approach, Dan also conjugated an oligosaccharide derived from heparin to gold nanoparticles this way. The oligosaccharide had previously been derivatised by Nina Azmi, so that it had a maleimide at its reducing end (thiol chemistry is orthogonal to the reactivity of other groups on this class of sugars, so particularly useful). The ligand shell of the nanoparticles, on which we have spent a lot of effort in the past, is resistant to ligand exchange by dithiothreitol (the present paper and here). So a quick reaction of the maleimide with dithiothreitol, and we have a thiol on the outside of the nanoparticle ligand shell, which in turn reacts with the maleimide on the oligosaccharide.
The ligand shell allows control of the stoichiometry of nanoparticle functionalisation, so Dan made 1:1 conjugates, which will be particularly useful for future functional studies.
We are excited by this work, because hitherto we have been restricted to non-covalent conjugation. A perpetual worry is that the exchange of the hexahistidine tagged protein on the nickel-NTA-nanoparticle for an endogenous protein with a histidine patch. We know there is real potential for this, because, when you run a cell lysate over a nickel-NTA column, there are plenty of untagged proteins that bind rather well. With covalent conjugation, this is one worry we can lay to rest.
In the case of proteins, cysteine can often be mutated to serine without affecting function and/or a cysteine can be introduced genetically as an N-terminal tag, again without affecting function. This allows rational chemical genetic covalent coupling of proteins to nanoparticles by Dan’s method, which provides scope for applications in functional work in cells. For sugars and nucleic acids, the thiol chemistry is, as noted above, completely orthogonal to the groups that occur naturally on these molecules, so nanoparticle conjugation can be designed very precisely and simply.
Update 2 October 2014 Our library has a deal with the Royal Society of Chemistry, which I was not aware of, which has allowed Dan’s paper to be Open Access, without recourse to lab funds – there were none for this purpose, since paying of AO depends on the funding organisation, etc. I guess the deal is along the lines that we get a certain number of OA articles because we subscribe to RCS journals. Another reason (others here) to publish with at least certain Learned Societies – they don’t double dip (pay for subscription, pay for OA) unlike many commercial, closed access publishers.