New techniques to diagnose sepsis only take a few hours

In the United States each year there are more than 1 million sepsis, more than 250,000 deaths - this is a disease that kills more people than breast cancer, prostate cancer and HIV combined. In this situation, researchers are still rushing to find therapies as well as their causes and efforts somewhat paid off when the research team from the University of California in San Diego (UCSD ) has obtained a diagnostic tool that allows bacteria to be identified in just a few hours, significantly shortening the diagnosis time which normally takes several days as current methods.

UCSD's method of exploiting 2 types of new technologies including microfluidic (micro-liquid, also known as lab-on-chip technology) and High Resolution Melt (HRM - high resolution molten heat analysis - a biochemistry Molecular learning allows the detection of bacteria through a trace that it leaves on the DNA sequence called " the melting curve" .

The team used microfluidic technology to detect all the DNA that is attacked by the bacteria in the DNA sample. The usual method is to select each DNA that they suspect is the cause of inflammation and obviously this method is ineffective, time consuming.

"Bacterial DNA is present on everything and infects everywhere," Stephanie Fraley said.

Stephanie Fraley, associate professor of biological engineering at UCSD, said: "Bacterial DNA is present in everything and infects everywhere, so trying to find the DNA associated with septicemia is like picking a needle. tank " . Fraley was awarded the Burroughs Welcome Fund Career Award for the technologies she developed for the treatment of sepsis and research today is also based on previous Fraley achievements.

Regarding HRM diagnostic method , the advantage of this method is that it can help determine genetic composition in a series of rapid and accurate test samples. In this study, the researchers classified the genotype at the same time with 20,000 reactions from 1 ml of blood given Listeria monocytogenes (food-borne infection) and Streptococcus pneumoniae (bacteria that cause inflammatory diseases through the road). respiratory and meningitis)

To do this, they first extracted all the DNA from the blood sample and put it into an electronic chip that allowed each piece of blood sample to reproduce on its own. Each compartment on the chip can only contain 20 picoliter! They used a special chemical mixture to get a very small amount of blood into the chip and the technique is also waiting to be patented. It is worth noting that in the USCSD experiment, it is normal for HRM analysis to be performed only after the completion of a molecular replication technique called chain polymerization or Polymerase Chain gene amplification reaction. Reaction (PCR - a technique for creating multiple copies of a DNA fragment without using living organisms like E. Coli or yeast). In this case, the researchers developed a computer algorithm that allowed them to bypass the PCR technique and perform automated HRM analysis.

Fraley said: "No one has ever analyzed many reactions at once in such a small scale. Most molecular tests on DNA are usually done at a larger scale and can only be searched for in one category. Bacteria every time, we analyze all the bacteria in the blood sample. This is a huge step forward. "

Fraley's current goal and team is to bring this diagnostic tool out of the lab.

Next, the researchers heated the chip with amplified DNA with a gradually increasing temperature of $ 0.2 C, slowly causing the DNA to melt. When double helix DNA splits, connections between DNA sequences are broken, revealing specific signs. To recover this signal, the team used a special type of fluorescent dye that could bind to double-stranded DNA, causing these signals to glow. When the chains are decayed, the fluorescence becomes weaker because the double-stranded DNA is no longer allowed to stick.

From these signs, they get what is called the "melting curve" of bacteria by microscopic observation. This curve was then analyzed by a machine learning algorithm that had previously been loaded with 37 different types of bacteria. The end result shows that this algorithm is much more efficient than traditional methods with an accuracy of up to 99%, while the method usually has a error rate of up to 23%.

Fraley's current goal and team is to bring this diagnostic tool out of the lab. In the future, researchers will seek to minimize the size of the system so that it can be conveniently used in clinics and clinical institutes as well as expand the ability to detect pathogens including fungi and viruses. and anti-antibiotic genes. In addition, they also need to carry out additional studies based on actual blood samples from patients. Fraley hopes the system will reach doctors in the next 5 years.