Aptamers are interesting beasts. In general, they’re nothing more than oligonucleotides (DNA or RNA) that have been screened/selected to recognize some particular target. Because the machinery of molecular biology lets you generate truly insane numbers of different oligos, you can screen vast landscapes for binding partners and amplify them once one turns up, even if it’s present in very small quantities. Some of these binding interactions are very tight and specific indeed, and a new paper goes into details about a particular aptamer that was found to bind the natural product ochratoxin A. 

That’s shown at right – there are several of these things, mycotoxins produced by many Aspergillus and some Penicillium strains. They’re common trace contaminant in many foods (very trace, fortunately), but larger amounts can lead to kidney damage. This may in fact be behind some puzzling regional hot spots of nephropathy, such as in some areas near the Danube. It’s also a probable carcinogen, although the evidence in humans is less definitive than in some animal models. At any rate, it’s bad stuff, and you only want to be exposed to minimal levels of it. Ochratoxin B has exactly the same structure, minus the chlorine atom.

Which makes it interesting that there’s a particular 36-mer, known as OBA36, which binds ochratoxin A (OTA) with about a 50 to 100 nM K_D but binds ochratoxin B (OTB) at least 100-fold less potently. What is it that makes it so selective? A good OTA-binding reagent could potentially be used in all sorts of detectors and diagnostic tests, but it would be worth knowing how it manages to do what it does, with only one chlorine atom’s difference to work with. OBA36 is rich in guanines, so it was suspected that there was a “G quadraplex” structure involved, but where, and to what end?

The new paper has NMR structural evidence that a G-quadraplex does indeed form when OBA36 (actually, a related 33-mer) binds to OTA, but that it’s not really the binding site per se. Instead, OTA binds right where the G-quadraplex starts turning into a double helix, and a set of hydrophobic interactions and pi-stacking interactions holds the ochratoxin molecule in place. . .along with a crucial halogen bond. OTB (the des-chloro) binds in an almost identical manner according to the NMR data, but it of course lacks the halogen bond, and that’s what makes the difference. That’s one of those interactions that we don’t take advantage of in this business as often as we probably should. Broadly speaking, halogen bonds are the pairing of an electrophilic region around a halogen atom with a corresponding nucleophilic region in the binding partner, but this has several different components when you start to unravel the specific mechanisms of attraction (electrostatics, charge-transfer, and more). It makes sense that a halogen bond would be the secret ingredient here (the chlorine points into the pyrimidine ring of a particular C reside in the aptamer), but it’s good to see this proven with real data. That single chlorine appears to contribute about a third of the favorable free energy change all by itself.

That quadraplex-duplex structure of the aptamer is itself held together by Mg and Na ions, but in a rather flexible form. Some of the variability in the assay numbers in the literature might well be due to the ionic strength of the solutions involved; those cations are key. The arrival of an ochratoxin molecule causes the aptamer to adopt a more defined structure, and ochratoxin A’s is obviously even more well defined than the B. The “quadraplex duplex” motif seen here is apparently a new aptamer binding motif all by itself, and you wonder if there are other halogen-containing targets that might also take advantage of this sort of thing, since that halogen-cytosine interaction can be so lucrative from a free-energy standpoint. Binding energy is where you find it!