How to quickly identify sepsis-causing bacteria – melt it down
When a patient is diagnosed with sepsis, a medical syndrome that kills more people than breast cancer, prostate cancer and HIV combined, it sets off a countdown for doctors to treat the infection and uncover the culprit causing the body's systems to shut down. However, identifying the exact pathogen causing the infection can take days with current procedures, which is time a terminally ill patient simply does not have. But hope could be on the horizon, as researchers from the University of California San Diego (UCSD) recently unveiled a diagnostic tool that enables blood-borne bacteria to be identified in a matter of hours.
The new method employs two emerging technologies – microfluidics (also known as lab-on-chip technology) and High Resolution Melt (HRM) – to identify bacteria through a unique DNA sequence-dependent fingerprint known as melting curve.
Sick of Ads?
More than 700 New Atlas Plus subscribers read our newsletter and website without ads.
Join them for just US$19 a year.More Information
With the former, researchers can see all the bacterial DNA that is present individually in a sample, instead of just singling out the one they think might be the cause of the inflammation, which is an inefficient way of dealing with a condition where speed is of the essence.
"Bacterial DNA is on everything and contamination is everywhere, so trying to find the ones associated with sepsis is like the proverbial search for the needle in the haystack," noted lead author Stephanie Fraley, an assistant professor of bioengineering at UCSD, back when she was awarded the Burroughs Wellcome Fund Career Award to develop engineering technologies for the treatment of sepsis. The new study builds on Fraley's work in this field.
As for HRM analysis, what makes it so advantageous is that it is able to determine the genetic make-up of large batches of samples quickly and accurately. In the case of this study, the researchers were able to genotype 20,000 simultaneous reactions from just one milliliter of blood that had been inoculated with the food-borne germ Listeria monocytogenes and with Streptococcus pneumoniae, a bacterium that is responsible for respiratory infections and meningitis.
To do this, the researchers first isolated all the DNA from the blood sample and placed it on a digital chip that allowed each piece to independently multiply in its own reaction. Each well contained only 20 picoliters – a single raindrop, by way of comparison, typically contains hundreds of thousands of picoliters – a feat that was achieved via a proprietary mix of chemicals. What is noteworthy about the UCSD experiment is that usually, HRM analysis is conducted after the completion of a molecular photocopying technique known as polymerase chain reaction. In this case, the researchers developed a machine algorithm that allowed them to skip this process and conduct the analysis automatically.
"Analyzing this many reactions at the same time at this small a scale had never been attempted before," says Fraley. "Most molecular tests look at DNA on a much larger scale and look for just one type of bacteria at a time. We analyze all the bacteria in a sample. This is a much more holistic approach."
Next, the researchers heated the chip with the amplified DNA in increments of 0.2 degrees Celsius, causing it to melt. What happens next is key to understanding what makes HRM analysis such a powerful genotyping technique, and why it could help advance our fight against sepsis.
As the DNA double-helix melts, the bonds holding together the DNA strands start to break down, revealing its unique signature. This is captured using a special kind of fluorescent dye that binds to double-stranded DNA, causing it to glow brightly. As the strands dissolve, the fluorescence becomes weaker since there is nothing to which the dye can adhere itself.
The engineers captured the bacteria's melting curves using a specially designed high-throughput microscope and then analyzed them with a machine learning algorithm that had previously been trained on 37 different types of bacteria. There was no contest: the algorithm outperformed traditional methods, which have an error rate of nearly 23 percent, by identifying bacteria strains with 99 percent accuracy.
The goal now for Fraley and her team is to take this diagnostic tool out of the lab and into doctors' offices. Though other researchers have attempted to solve the problem of sepsis by developing artificial spleens that can clean the blood as well as bacteria-repellant surface coatings for medical devices, there remains an urgent need for more efficient diagnostic tools.
Moving forward, the researchers will be working to shrink the size of the system to make it suitable for use in clinics and physicians' offices, and expanding the system's pathogen-detecting abilities to include fungi and viruses, as well as genes for antibiotic resistance. In addition, further studies will need to be conducted on actual patient samples.
Fraley is hopeful the system will be available to physicians in the next five years. "This has the potential to reach people near or at the point of care," she says, adding that with the right adjustments, there is also potential for it to be deployed in low-resources settings. "It's a simple and innovative approach."
The study was published in Nature Scientific Reports.