Teixobactin: NRPS from newly cultured bacterium

Last week, a paper published in Nature stormed through every major news outlet. I’ve never seen anything quite like it: I was getting calls and emails and no matter what news source I looked at, I found coverage of this paper. To be fair, I think this sort of thing happens on a fairly regular basis but this happened in my field so I felt a bit of the force of the media in full throttle. What was the paper? The paper comes from Kim Lewis' lab at Northeastern where he has used a culture-based approach to identify a rare bacterium, and from that bacteria to isolate a new antibacterial compound he calls teixobactin. The paper describes the use of an unconventional culturing technique (the “iCHIP”) to culture a rare bacterium from which he was able to isolate a potent antibiotic. This antibiotic kills a wide range of organisms both in-vitro and in a mouse model of infection and they were able to postulate a mechanism of action (inhibition of cell wall biosynthesis) and drug’s target molecule (lipidII/lipidIII). All in all this was a really beautiful story: it highlights the power of looking for new antibiotics in the uncultured majority of microbial life present in soil environments all around us. Furthermore, it has the fully integrated story including activity in a physiological environment (mouse model of infection). Its the sort of paper anyone in the field would be proud of. So whats my problem?

Theres two problems with this paper that are partially due to the immense current interest in the need for antibiotics and the media scrutiny that resulted: the idea of being “irresistible” and the implication, given by many of the news stories, that this compound will lead immediately to something in the clinic. The first issue, resistance, was raised in part because of the observation that in a serial dilution assay looking for natively resistant mutants, teixobactin, but not other antibiotics, did not select for resistant mutants. While this is a great finding, it has been taken WAY out of context! There will be resistance to ANY antibiotic whether through modification of the compound, efflux pumps, or the emergence of an alternative target. In a Nature paper a few years ago the Gerald Wright group found vancomycin resistance cassettes in DNA extracted from 30,000 year old permafrost sediments. Resistance mechanisms to this drug are far older than the drug’s discovery. All it took for the emergence of these genes to be a clinical problme was the consistent useage of Vancomycin.

The actual discussion section includes the following:

Polyprenyl-coupled cell envelope precursors, such as lipid II, are readily accessible on the outside of Gram-positive bacteria and represent an ‘Achilles heel’ for antibiotic attack. The target of teixobactin, the pyrophosphate-sugar moiety of these molecules, is highly conserved among eubacteria. The producer is a Gram-negative bacterium, and its outer membrane will protect it from re-entry of the compound. This suggests that the producer does not employ an alternative pathway for cell wall synthesis that would protect it from teixobactin, and which other bacteria could borrow. Resistance could eventually emerge from horizontal transmission of a resistance mechanism from some soil bacterium, and given the highly conserved teixobactin binding motif, this would likely take the form of an antibiotic modifying enzyme. However, although determinants coding for enzymes attacking frequently found antibiotics such as β-lactams or aminoglycosides are common, they are unknown for the rare vancomycin.

This makes the “irresistible” argument a bit more palatable. While concending that an antibiotic modification enzyme would render the drug ineffective, the other common classes of resistance are less likely to occur: efflux wouldn’t work because the target is exterior, and target modification would difficult because of its conservation. While these do seem to be compelling arguments for teixobactin having a lower native rate of resistance, it in no way invalidates the logic of antibiotic use and the inevitable spread of resistance once the modification genes have been identified.

The second issue, that this compound will lead to a clinically useful compound is still up in the air. As Kim lewis says in this interview: “I think we’re probably two years away from Phase I clinical trials and five years away from having the drug in the clinic.” That is a totally reasonable timeline for a promising compound. The news stories suggesting a sea change in antibiotics, therefore, need to wait in order to see if these compounds live up to their promise. I would be completely happy if this compound does live up to the expectation but the low success rates of clinical trials should urge caution.

Although I have been largely negative in the last few paragraphs, I think I should end by commending Kim Lewis and his team on a job well done. This is indeed a beautiful example of how new ways of tapping into the biodiversity of soil bacterium can yield structurally interesting and potent bioactive compounds. I hope teixobactin, or one of the 25 other compounds he found, succeed and I hope the attention being given to this field sparks a larger interest in the general effort that, of course, our lab is a part of.