June 11, 2013
Due in large part to its overuse, experts say that within the next decade or two, the world is at risk of running out of any effective antibiotics to treat bacterial infections. This is due to bacteria gaining resistance to antibiotics at an alarming rate. University of Illinois researchers, working under professors Charles Schroeder and Paul Kenis, are developing a potential solution to this problem.
At the heart of the issue are physicians’ practices of prescribing several combinations of antibiotics hoping that one of them works.
“What doctors typically do is prescribe a combination of four or five antibiotics,” Ph.D. student Arnab Mukherjee said. “Maybe one of the five work, but because the bacteria is exposed to the other four antibiotics, they develop resistance against it over time.”
For Mukherjee, the pursuit is personal. A graduate of the Indian Institute of Technology, Mukherjee watched as a colleague of one of his professor’s was infected with flesh-eating bacteria after suffering a seemingly harmless shoe-bite. Mukherjee says that doctors treated the problem with as many as 15 different antibiotics. They eventually resorted to flushing her body internally with low doses of bleach, which, along with the infection, led to her death.
Inspired by that event, the potential solution came to Mukherjee three years ago after listening to a lecture from Kenis.
“Professor Kenis was talking about microfluidic chips that he designed to crystalize proteins as well as pharmaceuticals,” Mukherjee recalled. “ I thought that we could adapt these chips to study antibiotic resistance.”
Mukherjee believes this method is the most ideal way of tackling this problem given the fact the problem is already at a very acute stage and it can take between 5-10 years for the FDA to approve new drugs on the market.
While Mukherjee proposed the idea, the research project is a collaborative effort. The Kenis lab provides the microfluidics expertise, while for the Schroeder lab contributes microbiological and bioengineering knowledge. The primary research goal is to utilize the chips to screen a patient’s sample against antibiotics and quickly provide information on what drugs or combination of drugs would work best against the infection.
“The current technology takes several days to determine the best antibiotic to use against a particular infection,” Mukherjee said. “This would cut it down to about three hours.
Ritika Mohan, a Ph.D. student in chemical and biological engineering, along with undergraduates Emre Sevgan and Jaebum Lee, has conducted much of the research in the area, which is now t the later stages of the proof-in-concept phase. Mohan has presented this work at several meetings. Her poster entitled “A Multiplexed Microfluidic Platform for Antibiotic Susceptibility Screening” at the Institute of Biological Engineering Symposium in Indianapolis received first prize.
The technology utilizes microfluidics where the fluids move at a very low flow rate through a channel similar to the width of a human hair. One can fill these chips with the patient’s sample and antibiotics. The sample and antibiotics are then allowed to mix and antibiotic susceptibility information is available within hours.
In the eventual clinical application, dyes would stain the living bacteria green and allow researchers to use microscopy to readily determine how many bacteria are present. If a chamber in the microfluidic chip is heavily “green,” the clinician would know the antibiotic is not working. However, if the green bacteria dwindle in numbers after antibiotic treatment, the antibiotic is most likely effective. Once the effective antibiotic is discovered, it can then be administered to the patient.
“In our case, it’s not important to know what the bacterium is,” Kenis said. “The clinical way of treatment as it currently exists is to know first exactly what the bacteria is, then to treat the patient with a certain antibiotic based on prior experience. In our case, we just test the infection against let’s say 10 or 20 types of antibiotic mixtures and concentrations. We see what works and administer that.”
The ultimate goal is to provide a device that allows the clinician to take a small sample from a patient (e.g., blood or saliva) and test efficacy of the antibiotics prior to prescribing an antibiotic dosing regimen to a patient. Thus, the clinicians prescribe only the antibiotic or an antibiotic combination that is known to be lethal for the pathogen causing an infection.
The device requires less than a microliter of blood or saliva as opposed to several milliliters, which is another advantage it has over conventional methods for susceptibility testing. Each one costs less than a dollar apiece and its current facilities can be expected to produce about 100 per week. The group also plans to write for a Gates Foundation grant to take the devices to developing countries.
The project is still in the proof-of-principle stages, testing antibiotics against common pathogens. While the early stages of the project used E.coli as the model pathogen, Mohan is currently using Pseudomonas aeruginosa, which is responsible for about 20 percent infections acquired in the hospital, and with Klebsiella pneumoniae.
“Our next step is to screen real clinical samples (e.g.,saliva) after they are mixed with appropriate optical dyes,” Mohan said. “If that works out, we should be able to use time-lapsed fluorescence microscopy with actual clinical samples.”
At the same time, Mohan is studying the affects of antibiotics on polymicrobial infections, those caused by more than one bacterium. She says that goals are twofold – to study the interaction between the pathogens as well as how the antibiotics affect those interactions.
“That’s a big deal because there is not a lot known in this area in microbiological labs or in clinics,” Mohan said. “Microfluidics is great tool to study polymicrobial infections because it enables single cell resolution, allowing us to visualize and quantify single cell dynamics in mixed populations over extended periods of time. Conventional techniques such as the “streak plate methods” are end-point assays over a given time.”
“It’s simple to use,” Kenis said. “It’s a matter of technology integration to get to the next step.”