The remarkable transformation in the control of infectious diseases by antibiotics is one of the glorious stories in microbiology. But now, almost inseparable from their discovery and application, is its nasty sequel, the rapid evolution of antibiotic resistance. We continually read reports of the dire public health situation that we may well confront in a few decades, the so-called "post antibiotic era." So, scientists and policy makers worldwide strategize about how to avoid this catastrophe. It is likely that halting the current trend will require efforts along multiple fronts. Among them are the enforcement of strict policies regarding the use of antibiotics and continued efforts to discover and develop new ones.
As important is a growing sense that current antibiotic therapies represent the ecological equivalent of "scorched earth" affecting pathogens and commensals. We must stop lobbing grenades at our friends as well as our foes. As a consequence, many microbiologists are trying to develop therapies that better maintain the ecological balance of the microbial communities of the host. There have been some very exciting developments relating to this latter concept in recent years and I here describe some of these as well as proffer my opinions on the subject.
In my view, some of the key questions in the field of antibiotics that remain largely unanswered relate to the ecological function that these molecules play in natural settings. The majority of the antibiotics used in clinical settings today are small molecules produced by microbes, most notably members of the Actinobacteria. We know that in the large fermentation vats of the pharmaceutical industry, heavily mutagenized antibiotic-producing strains are grown in rich media to yield the largest possible amount of antibiotics. But what is largely unknown is just how much antibiotic do these microbes make in their usual habitats. In nature, do they make enough antibiotics to kill their neighbors? Add to that the fact that we now know that among soil dwellers, many of the Actinobacteria present are resistant to the antibiotics produced by their neighbors. So, in that context, are these compounds actually used as killing agents? For decades, Julian Davies (Univ. of British Columbia) has been promulgating the idea that antibiotics act as signals to effect physiological changes in neighboring cells. To expand this thought, antibiotics could easily be multifunctional, killing nearby cells at high local concentrations but serving as signals to cells that sense only lower concentrations further away. There is a term for this, hormesis. Continuing to address these questions experimentally will lead to a better understanding of the ecological roles that these fascinating molecules play in their natural settings. There are, however, some instances where I think we have already gotten some clear indications that these substances indeed play an antimicrobial role.
In the mutualistic symbioses of insects with Actinobacteria, small molecules are used to control infections. My favorite story along these lines is that of the leaf-cutting ants, which was previously covered in this blog. Briefly, leaf-cutting ants are remarkable farmers. They collect leaves for the growth of fungi they cultivate in their nests, which they then eat. These fungi are susceptible to infection by different, pathogenic fungi. The ants manage to keep these pathogens in check by a multi-pronged approach: They thoroughly clean their "fungus gardens," plus, they harbor on their bodies Actinobacteria that produce antifungal compounds that kill the pathogenic fungus but leave their cultivated fungus intact (Figs. 2A & 2B). This is an ecological application of antimicrobial substances. Perhaps there is an important lesson here that we as humans should learn from the ants: target the undesirable microbes with specific antibiotic-producing microbes and leave the rest of our associated microbial communities intact.
But maybe we are a lot more like the leaf-cutting ants than we suspect. As the rate of discovery of new antibiotics from soil bacteria slowed down, people began to search in new and exotic locations, from the exoskeletons of insects to marine sponges and everything in between. Curiously, it was not until recently that the search for antibiotic producing bacteria among the members of the human microbiota began in earnest. The results obtained to date are, I think, extremely exciting. It is now becoming quite apparent that, akin to the ants, we humans harbor beneficial bacteria that produce antibiotics that keep potential pathogens at bay. Two recent papers illustrate this point beautifully.
The first of these papers, published in 2014, reported work by Michael Fischbach (UCSF) and collaborators. In this study the authors scanned the genomes of nearly 2500 human-associated bacteria for the presence of "small-molecule Biosynthetic Gene Clusters" (BGCs). These are gene clusters predicted to encode the proteins required for the production of the types of molecules that often have antibiotic activity. They then compared commonly found BGCs against 753 metagenomic samples from the NIH Human Microbiome Project. It was remarkable finding just how widespread were BGCs in the genomes of the human microbiota. One particularly prevalent and varied class of BGCs encoded antibiotics known as thiopeptides. They then isolated and characterized one novel thiopeptide. They called it lactocillin, because it was produced by a strain of Lactobacillus gasseri, one of the four Lactobacillus species commonly found to dominate the vaginal community. The authors showed that lactocillin was active against numerous human pathogens and commensals. However, they did not address the question of whether carriage of a lactocillin-producing strain by the host kept bacterial pathogens at bay. Such an indication came from a different system presented in the second paper.
The second paper reports work from a group led by Andreas Peschel at the Univ. of Tübingen, Germany. The authors started out by screening a large collection of nasal Staphylococcus isolates for antibiotic activity against a kin of theirs, the human opportunistic pathogen Staphylococcus aureus. They were interested in finding strains that antagonized S. aureus, because about 30% of humans carry S. aureus in their nasal passages and this carriage predisposes individuals to invasive infection by this opportunistic pathogen. Importantly, eradication of carriage with antibiotics reduces this predisposition. In short, they were interested in other Staphylococcus strains that might prevent S. aureus nasal colonization. Their screen yielded one strain of Staphylococcus lugdunensis that inhibited the growth of S. aureus in vitro by producing an antibiotic substance.
I particularly liked how the authors determined the genes responsible for the production of the antibiotic; they did it the old-fashioned way. They carried out transposon mutagenesis of the S. lugdunensis strain, screened for one mutant that lost antibiotic production, and sequenced the transposon insertion site. This way they identified a BGC (biosynthetic gene cluster, remember?) predicted to encode the proteins needed for the production of a "non-ribosomal peptide" type of antibiotic. Now they knew what they were looking for and, using typical activity-guided purification and characterization, they got in hand a novel antimicrobial they called lugdunin (Fig. 3). It turns out that lugdunin is active against many different Gram-positive bacteria.
Had they stopped here they would have had a very nice story already. But what they did next with the lugdunin-producing strain is, in my opinion, what makes this story so exciting. They showed that S. lugdunensis interferes with S. aureus colonization in vivo using a nasal rat model. And to add icing on the cake, they did a preliminary epidemiological study determining the distribution of S. aureus and S. lugdunensis in hospitalized patients. While the number of patients was relatively low (187 in total) they obtained good evidence that the presence of S. lugdunensis – through the production of lugdunin – prevents human nasal colonization by S. aureus. This is all extremely "neat" and it could have important implications on the way we apply antibiotic therapies in the future. I will not be surprised if within a few years we begin to whiff S. lugdunensis up our noses. This may well encourage an ecological balance, where S. aureus colonization is reduced without severe disruption of our nasal microbiota.
Imagine that; who would have thought that humans can harbor mutualistic bacteria that produce antibiotics involved in keeping our microbiota in ecological balance? I tell you who… Way back in 1976, when I was a graduate student, I came across a wonderful little paper that described the identification of a "New Family of Low Molecular Weight Antibiotics from Enterobacteria". Carlos Asencio and colleagues at the Universidad Autónoma de Madrid had discovered the microcins (and lugdunin is a microcin indeed) based on the ecological premise that the species composition of the microbiota of the human intestinal tract is defined in part via the production of antibiotic substances. In reading the lugdunin paper of 2016, I marvel at the foresight that Asensio and colleagues showed forty years ago. As we look ahead, it’s never too late to look back at what our forerunners discovered.