Bullvalene is a pretty strange molecule. It was first reported in 1963 from the lab of William Von Eggers Doering at Yale, and it was made with just that strangeness in mind. The name, by the way, comes from Doering’s nickname of “Bull”, in case you were wondering. The molecule looks a bit odd, but symmetrical and seemingly innocuous. But that arrangement of double bonds and the cyclopropane ring allow for very easy Cope rearrangements that end up moving the cyclopropane part and the alkenes around. You can draw those things all day long, and Doering and co-workers predicted that in fact bullvalene would be a “fluxional” molecule without a fixed structure at room temperature.
That’s exactly right. The proton and NMR spectra of the compound show one peak each at room temperature, which from the drawing shown just shouldn’t be right. I mean, the cyclopropyl should be different from the other alkane end, and you have those alkene carbons/hydrogens in the middle – if you asked a student who’s never heard of the stuff to predict the NMR spectra, they’d predict four carbon peaks and four proton peak, most likely. But you don’t get that until you cool the molecule to really low temperatures. Sitting on the bench, it’s rapidly interconverting, and if you could stick some sort of colored dot on each carbon to tell them apart, you’d find that there are over 1.2 million possible tautomeric structures. Every carbon gets to be part of the cyclopropane ring, the terminal alkane at the other end, and the alkene carbons, and every carbon gets to be every one of those, over and over in a ceaseless high-speed dance.
That fluxional behavior is more common than most of us organic chemists think about most of the time (what with ring conformers, amide rotamers, and more), but bullvalene is the crowning example. And it’s been a curiosity of the chemical literature for decades now, and until today I don’t think I’ve ever seen it put to any practical use (but that’s because I missed this one from 2013). But here’s a new paper in PNAS (earlier ChemRxiv version here) that uses bullvalene as the centerpiece of an antibiotic structure, and your first thought is why anyone would do such a thing, right?
They’re looking at vancomycin, which is famously an antibiotic of last resort for difficult resistant Gram-positive organisms. And that’s because its mechanism makes resistance much harder to develop – instead of messing up some enzyme’s active site, vancomycin binds to the D-alanine-D-alanine motif that’s used in Gram-positive bacterial cell walls, and that binding prevents the cell-wall cross-linking that the bacteria need in order for that wall to function property. But if you give them enough time, bacteria will stumble into ways to overcome most anything, and there are indeed vancomycin-resistant organisms (and indeed you do not want to be infected by them, it should go without saying). Instead of D-Ala-D-Ala at the end of their cell wall peptides, they have D-Ala-D-lactate, which allows the crosslinking to take place but keeps vancomycin from binding.
There have been a number of ingenious efforts to produce vancomycin derivatives and combinations that can get around this problem. Among these have been the production of various vancomycin dimers, and this new paper extends that idea to putting a bullvalene in the middle of the linker between two vancomycins. This builds off a recent synthetic route that makes it much easier to produce disubstituted bullvalenes, otherwise this would have been a terrible slog for sure. The disubstitution cuts things down to fifteen positional isomers at the bullvalene. The team here prepared several such linked dimers, as well as non-fluxional control dimers, and tested them against bacteria.
These tests were both for antibiotic activity (which many of the compounds showed) but for the ability to evade the bacterial attempts at resistance. You do that by growing these bacteria under gradually increasing nonlethal concentrations of the antibiotic, slowly raising the bar and challenging the organisms to come up with a way out of the situation. Under these conditions, the control-linked dimers did indeed bring on resistant bacterial strains that could tolerate what should have been lethal challenges, but the bacteria could not get around the bullvalene-linked dimers, and the hypothesis is because these presented an ever-shifting target. But it’s tricky to establish what’s going on in detail, because the vancomycin dimers almost certainly work through mechanisms that “regular” vancomycin doesn’t use (it’s not just D-Ala-D-Ala binding). Since you’re getting up to fifteen different structures simultaneously here, it won’t be easy!
As the authors say (correctly) “the potential benefits of molecular shapeshifting have, until now, been overlooked”. It remains to be seen what the other properties of these dimer molecules are like and whether they can be developed into drugs, but the general principle will surely get more attention, and it can be applied to other antibiotics as well. The resistance problem is a serious one and needs all the ideas we can throw at it. Bullvalene, who’da thought. . .
