Bacteria have evolved many ways to defend themselves against antibiotics, resulting in devastating infections resistant to many antibiotics. Dr. Michael Mahan and colleagues from University of California Santa Barbara have discovered an important and novel way bacteria evade our defenses.
Now and then a new article grabs my attention with an “aha” or a “eureka” moment, presenting a game changing idea. Despite a rather dense title, this new study left me with that reaction, and wondering why no one had thought of this before. The authors tried to examine why some people fail antibiotic therapy even though the bacteria isolated from them appear susceptible (or sensitive) to that antibiotic on traditional testing. They found a host-dependent mechanism that would not be detected by the usual tests.
Perspective
We’ve become too reliant on “standard” tests, letting that frame our thinking. This is reminiscent, for example, of the recently debunked teaching that urine is normally sterile. In a study earlier this year, Loyola researchers observed that standard urine cultures only detect specific organisms, and are unable to detect bacteria with special growth requirements. Thus, clinicians are likely missing infections because of the limitations of the traditional culture process…and we are trained to believe the negative results, and that there is no infection, rather than considering the limitations of that test.
A similar pattern happened with pneumonia. Until the early 1980’s, we were unaware that “atypical” bacteria like Chlamydia could routinely cause pneumonia. I learned this first hand while conducting a clinical trial in which we sent special cultures to Dr. Thomas Grayston at University of Washington, and found a number of patients' specimens were positive for mycoplasma or chlamydia. We can’t routinely culture for these fastidious organisms; standard treatment guidelines for pneumonia now recommend antibiotics for such atypical organisms.
As I noted in “Ticked Off: What We Don't Know About Lyme Disease,” we are similarly limited in our ability to diagnose and treat Lyme disease. We have no good assays to diagnose or monitor response to therapy, and there is increasing thought that there may be intracellular or persisting forms of the Borrelia that are not killed by traditional antibiotic courses.
Antibiotic resistance depends on different factors in the host
Currently, all antibiotic susceptibility testing is standardized, using bacterial growth on Mueller–Hinton Broth (MHB). Instead of using only MHB, the investigators in Mahan’s lab also used a culture medium called LPM, which more closely mirrors intracellular conditions, where some bacteria, like Salmonella, replicate. They found that the bacteria were much more resistant to some classes of antibiotics tested in the low pH, intracellular-like media, even though they looked susceptible by traditional testing. Further, resistance could be promoted by use of the wrong antibiotic, even if it was one that looked like it “should” work.
Mahan’s team found a similar pattern using Yersinia pseudotuberculosis, which does not multiply within the cells, but lives extracellularly in the intestines. Again, resistance could be induced in cultures that mimicked that environment.
As Mahan explained, there is a need “for lab drug sensitivity testing to incorporate media that mimic the specific biochemical environments that trigger resistance in the body. If these additional components were added, we’d get a more accurate reflection of what happens when a patient is treated with a particular drug.”
Mahan explains his study very well in this video:
For those who want more details…if not, skip ahead
Instead of using MHB and a relatively neutral pH of 7.2, the researchers tried to replicate an intracellular environment. For example, Salmonella can even reproduce within macrophages, a type of white blood cell important in host defenses, in a microenvironment that can be mimicked by an acidic culture medium (pH 5.5) that is low in phosphate and magnesium (LPM). So the researchers compared the growth of Salmonella enterica Typhimurium in MHB to that in LPM media in the presence of different antibiotics. They found that the growth of the Salmonella in the LPM media was linked to high resistance to two antibiotics of last resort, polymyxin and colistin. The efficacy of some other antibiotics (e.g., tetracycline) was unaffected.
The scientists then screened Yersinia pseudotuberculosis, a Gram-negative extracellular pathogen, for resistance at different pH and using LPM media, finding this bacteria acquired resistance to different antibiotics, and under markedly different conditions.
More intriguingly, when the bacteria were transferred from MHB to acidic LPM media, Salmonella rapidly became antibiotic resistant. When transferred back to neutral MHB, the bacteria rapidly reverted to susceptible again. This reversibility under conditions that simulated intracellular host environment was termed TIVAR, transient in vivo antibiotic resistance. Treatment of TIVAR+ bacteria with certain antibiotics might induce resistance that then becomes permanent through subsequent mutations.
TIVAR overview. (A) The standard 7–10 day antibiotic treatment regimen is sufficient to clear most... [+]
Further testing showed that drug resistance could also be induced in iron-rich environments, which might be a factor in the emergence of resistance from agricultural uses of antibiotics or in patients with certain problems with iron metabolism.
Why this study is so important
As Dr. Mahan aptly notes, “We are not petri plates… and the host environment triggers changes in the bugs so they become resistant to high levels of certain antibiotics...explaining why the standard clinical tests do not always correlate with patient outcome.”
Besides showing a possible mechanism for why antibiotics that appear susceptible by standard lab testing (in vitro) might not work in people (in vivo), this study also suggests that traditional testing might exclude effective antibiotics from consideration. In another recent study, for example, Dr. Victor Nizet and colleagues at the University of California San Diego showed that azithromycin, an antibiotic commonly used for respiratory tract infections, but never for serious Gram negative infections, was effective in killing the superbugs, carbapenem-resistant isolates of Pseudomonas aeruginosa, Klebsiella pneumoniae, and Acinetobacter baumannii, first using cell culture media and then in mice.
Mahan notes the need “for lab drug sensitivity testing to incorporate media that mimic the specific biochemical environments that trigger resistance in the body.” For example, using LPM and cell culture media could give us a much more accurate prediction of whether antibiotics would work in people.
This new model helps explain why patients might not be responding to antibiotics the way we expect them to. When I encounter such a situation, I normally ask myself: Is this the wrong drug? Wrong dose? Is there an undrained focus the antibiotic is not reaching? Instead, the problem may be that the traditional way we test antibiotic sensitivities is too limited.
This type of testing is also exciting as it might produce a rapid way of finding new drugs for the growing problem of drug-resistant infections. Mahan explains that pharmaceutical companies already hold large 'chemical libraries' containing “millions of chemical compounds that were excluded as antibiotics on petri plates (MHB); they may work well to fight infections...but they were never tested since standard petri plates ruled them out.” As he then enthusiastically described, if these compounds were screened using the other media, as his and Nizet’s group did, more effective compounds might be relatively readily found.
This paradigm shift requires consideration of intracellular conditions or microenvironments within the body and presents interesting new challenges—and opportunities—for improving antibiotic susceptibility testing, for guiding clinicians as they treat patients, and, hopefully, for reducing the emergence of resistance.
Publication: Host-dependent Induction of Transient Antibiotic Resistance: A Prelude to Treatment Failure. Jessica Z. Kubicek-Sutherland et al. EBioMedicine (2015) DOI: 10.1016/j.ebiom.2015.08.012
