Here’s a new paper with what might be an interesting approach to Lyme disease (and here’s Nature on this story). As it stands, Lyme is treated with a pretty stiff course of some pretty broad-spectrum antibiotics which naturally go on to disrupt a patient’s own gut microbiome. Antibiotics like that are extremely useful when you’re not sure what’s causing an infection, but for Lyme we know the culprit for sure: it’s the spirochete Borreliella burgdorferi (and perhaps some other very closely related species – fine species distinctions among bacteria are sometimes rather difficult to make). So this is one of those cases where a more selective agent could really be useful.
The authors conducted a screen in soil actinomycetes, which as they note are a pretty well-studied source of antibiotics – but not so much for really selective ones, because that’s not where the focus has been, historically. And they uncovered a compound that’s been known since the 1950s, hygromycin A (also known as totomycin). To the best of my knowledge, it’s never been developed for human use, because it was not seen to be especially potent against panels of common disease organisms. But it does hit B. burgdorferi and several other spirochetes, interestingly, while having much lower activity against common gut bacteria.
Table 1 of the paper shows the minimum inhibitory concentrations for it and three common antibiotics given for Lyme (doxycycline, amoxicillin, and ceftriaxone) against a good-sized panel of different bacteria. The first drug is a tetracycline, the second is an aminopenicillin, and the third is a later-generation cephalosporin. Hygromycin A is none of the above: broadly speaking, it’s in the aminoglycoside class, although it looks rather different than the ones in common use. As you look through the table, you can get a quick lesson in how much fun it is to prescribe antibiotics. All four drugs look to be potent against various Borreliella species (MIC values below one micromolar). But hygromycin looks very unimpressive against the whole opportunistic-pathogen list, which is why it was never followed up on all that much. It’s no good against (for example) Clostridium tertium KLE 2303, and neither is amoxicillin. But if you’re being infected with Clostridium perfringens KLE 2326, amoxicillin will work again. Your best bet of the four against Shigella sonnei ATCC 25931 is definitely ceftriaxone, but against Staphylococcus aureus HG003 you’re a lot better off with doxycycline (that one’s quite a bit better than ceftriaxone for that particular organism, which is interesting because the latter drug is a pretty general sledgehammer). As an example, ceftriaxone is the only one of the four that shows real activity against Salmonella typhimurium LT2 ATCC 19585. But none of them are any good against Pseudomonas aeruginosa PAO1, which I remember from my own antibiotic discovery days as a particularly nasty organism all over.
When you look at the gut bacteria, the difference between hygromycin and the others really does stand out. Hygromycin really doesn’t hit any of the 16 common species on that list at all. Amoxicillin is second-best at sparing the gut, although it will take a few of the ones on the list out, while doxycycline and ceftriaxone go through the gut microbiota like Tamerlane’s armies went through Asia. As we’ve realized more about the importance of the gut microbiota over the years, a profile like this has become more desirable. But how does it achieve that? Being inactive against gut bacteria is easy enough – I’ve made plenty of compounds like that – but having one actually be active against some other pathogen at the same time, that’s interesting.
The authors still aren’t quite sure themselves. Most aminoglycoside antibiotics exert most of their effect through targeting the bacterial ribosome and interfering with protein synthesis, and hygromycin A does that as well. And it’s hard to see how that can be such a selective mode of action. In fact, it seems to bind to the ribosomes of other bacterial species just fine, including species that it’s completely ineffective against as an antibiotic. The best guess is that it comes down to transport. Bacteria as a whole are annoyingly efficient at putting up barriers to extraneous substances – they have various sorts of membranes (nastily multilayered in the case of the gram-negative organisms, weirdly thick and gooey in the case of the Mycobacteria), and these are lined with a battery of efflux-pump proteins. These are the surly bar bouncers of the membrane protein world: they spend their days finding every molecule they don’t have on their lists and squirting them back outside, and that includes a long list of antibiotics. At the same time, there’s a long list of specific transporter proteins waving particular molecules in and rejecting everything else. Experiments that produced hygromycin-resistant bacteria (which was not easy, a good sign) suggest that it’s one of these transporters (bmpDEFG, more peculiar to spirochetes) that’s bringing the compound in, while other bacteria reject it.
The paper goes on to show that the compound is effective in a mouse model of B. bergdorfii infection, with the animals eating hygromycin-laced baits. Higher doses did not show any particular safety signals, and analysis of their gut bacteria showed them largely unaffected. Human cells show no toxic effects as well. The authors suggest hygromycin as a good candidate for clinical studies (and they make an excellent case), and go further to suggest that spreading the compound in baits might be able to eradicate the disease in the wild. That’s an ambitious goal, but worth studying as well. Good luck to all!
