If you look for them, you'll find them—bacteria that can feed on antibiotics, that is. Here, looking means to make an enrichment culture using a medium whose single carbon source is an antibiotic. Add a soil sample with its resident bacteria and look to see who, if anyone, will grow. As reported in a recent issue of Science, researchers collected soil samples from 11 environments, including some very unlikely to have had any exposure to human-made antibiotics. They tested each soil sample against 18 antibiotics, some natural and some human-made. There were no zero scores.
Every soil sample had bacteria able to eat at least one antibiotic, and bacteria from at least one soil sample were able to grow on every antibiotic. More than half of these talented organisms belong to the orders Burkholderiales and Pseudomonadales, groups with large (6-10 MB) genomes that encode an extensive metabolic repertoire. The majority of the isolates were resistant to the antibiotic used in their enrichment culture and some to others as well. Overall, 11 different orders were represented, so they cannot be relegated to some obscure taxonomic corner.
What to make of this? Obviously, it adds to the repertoire of bacterial biochemical talents. Other than some plastics, there is hardly any organic compound that cannot be chewed up by some bacterium (and we aren’t entirely convinced about the plastics, either—just give them a little more time). But does this finding matter for our health concerns about increasing antibiotic resistance? The concentration of antibiotics in soils is said to be very low even in soils that have not been treated with antibiotic-containing material, such as cow manure. It may not be so simple as all this because in the heterogeneous environment of soils, the local concentration of an antibiotic may be quite high, especially around the organisms that produce them (actual numbers seem hard to find in the literature, so would be grateful to readers who point them out to us). But if antibiotics are not present in significant amounts, being able to live off them would not be a pertinent skill.
Is "antibiotophagia" just a microbiological curiosity or does it matter in the evolution of drug resistance? Is this one of our Talmudic Questions?











It's an interesting find because there are bacteria that can metabolize man-made antibiotics. Does this suggest that metabolic processes in bacteria have evolved to be relatively non-specific? After all, Man-made antibiotics are, for the most part, analogs of 'natural' molecules.
From what I know about biochemistry, it would depend on the molecule, so perhaps we can use this general statement.
Posted by: ecoli | April 14, 2008 at 09:54 PM
I think it reflects something else, which is the degree to which researchers isolate themselves as they specialize. People who study antibiotic mechanisms and resistance often come from the med micro community, which rarely if ever pays attention to the microbial physiology and ecology community (I'm sure that there are many exceptions to this overgeneralization). Certainly when I worked on staph toxins, I paid no attention to studies on bioremediation and biodegradation, unless forced to in a class. However, if I had, I would have recognized that bacteria in nature are wondrously capable of extracting maximal value from potential substrates.
So I guess my point is that if someone were to synthesize the insights of these two areas (i.e. resistance happens and carbon sources get utilized) they would arrive at the notion that organisms in the soil should utilize antibiotics as carbon sources. It is certainly still an interesting finding that it is widespread in soil.
Posted by: Paul Orwin | April 14, 2008 at 12:26 PM
Please note, ecoli's blog URL is
http://blogs.scienceforums.net/ecoli
To comment on the comments by ecoli and Paul, it's of course been known for ages that bacteria can metabolize antibiotics (think for example of beta-lactamases). What is novel here is that bacteria can grow on products of such degradation. In truth this is not terribly surprising, given the bacterial metabolic repertoire, but it had not been appreciated before. And Paul, I also doubt that this has any clinical relevance.
Posted by: elio schaechter | April 14, 2008 at 08:18 AM
I agree with Paul Orwin above me. Why are we surprised that microbes can naturally metabolize antibiotics, when they can evolve resistance so quickly in vivo?
(I reiterate this in my own blog).
Posted by: ecoli | April 13, 2008 at 07:41 AM
The authors put very stringent restrictions on their search for antibiotiphages. Requiring them to use the chemicals as sole C source seems harsh. I'll bet that pairs (or trios) of bugs, one to break the first tough chemical bond, others to use the breakdown products, would give even higher frequencies of utilization. It's a very interesting study, isn't it!
-- stan zahler
Posted by: stan zahler | April 11, 2008 at 06:15 PM
I was surprised that this finding was surprising! I thought it was fairly clear that in the environment, there are lots of bacteria making antibiotics, and others that can break them down and/or resist them. We already know that various resistance mechanisms evolved in the soil and are still there, in the bacteria that make the drugs, and in others through HGT! I'm glad to see data that supports the idea, and I think the extent of antibiotiphagia (nice word!) is perhaps surprising to many.
I can't see any way that this doesn't affect clinical use of antibiotics. First, we know that there is plenty of movement of bacteria between soil and host organisms. Second, we know there is plenty of gene transfer going on in both environments. Therefore, there is a very high probability that within a soil microbial population, there is a potential pathogen (Burk and Pseudo being pretty strong contenders) that eats antibiotics for breakfast - so when the new antibiotic is introduced in the clinic, it just starts the process of enriching the host associated population for this strain. It also makes me think a lot about the "futility" of antibiotic development - we'll probably never develop something that bacteria can't resist/eat. On the bright side, it means lots of potential work for researchers!!
Posted by: Paul Orwin | April 11, 2008 at 09:47 AM