Chris Lock from Sciex explains how better diagnostics can help halt the potential pandemic of antimicrobial resistance.
In the medieval world, bubonic plague was a constant threat. Although most of us associate the disease with the Black Death of the 1300s – which wiped out up to two-thirds of all Europeans in four years and caused similar havoc around the world – plague persists in our modern world. Each year, the World Health Organization estimates that various forms of plague still infect several thousand people.
Better hygiene and pest control helped to eliminate the massive pandemics that periodically swept across the world. Fortunately, today, even those people unlucky enough to become infected with Yersinia pestis can readily be treated with a 10 to 14-day course of antibiotics such as doxycycline, gentamicin and ciprofloxacin.
While the discovery of antibiotics and their near-miraculous healing powers remains one of the biggest accomplishments of modern medicine, the rise of antimicrobial resistance (AMR) means that these drugs may soon be nearly useless.
‘Even the best drug will become obsolete if it is used improperly’
In 2019, antibiotic resistance directly caused the deaths of an estimated 1.27m people around the world and contributed to the deaths of an additional 3.68m. On its own, methicillin-resistant Staphylococus aureus (MRSA), which can result in sepsis and ventilator-associated pneumonia, caused more than 100,000 deaths worldwide in 2019.
While creating incentives to spur the development of new antimicrobial agents is one way to combat the rise in AMR, even the best drug will become obsolete if it is used improperly.
This challenge drives scientists to work towards reducing inappropriate antibiotic use and promoting more targeted therapies. These methods are being designed to rely on cutting-edge computational tools as much as biology and chemistry. By bringing these fields together, we can start to tackle the global AMR crisis.
A new approach to old diagnostics
Since the dawn of microbiology, scientists have identified bacteria by growing them in culture. Even under the best circumstances, it takes 12 to 24 hours to generate enough cells for an identification. Some types of bacteria take weeks, or even months, to grow. When faced with severe infections, that is time doctors do not have. They need an answer in minutes, not days.
Because they cannot afford to wait, doctors begin by prescribing broad-spectrum antibiotics that can kill a wide variety of bacterial species, and only switch to targeted narrow-spectrum therapies after the laboratory makes the identification. This technique is no doubt lifesaving, but it also perpetuates the development of AMR.
To tackle this, researchers around the world have been working to invent better diagnostic tools that can enable bacterial identification in a timely manner. Innovative scientists, including researchers in France, are developing an approach that can identify not only bacterial species, but also the types of antibiotics to which they remain sensitive.
The researchers’ strategy relies on mass spectrometry (MS), which separates ionised molecules based on size and mass. Each molecule reveals a unique spectrum of peaks, and scientists can also use liquid chromatography to purify samples before MS to eliminate spectral ‘noise’.
By comparing the results in a sample to a database of known spectra, researchers can identify nearly every compound. Since different bacterial species contain different hallmark chemicals, researchers can use the molecules identified via MS to determine which bacteria are present.
Prof Jerome Lemoine of the Institute of Analytical Sciences and Prof François Vandenesch of the Hospices Civils de Lyon, Hôpital de la Croix Rousse, both in Lyon, France, have been working together to develop a new MS-based analytical method to rapidly identify bacteria from clinical samples donated by hospitals for research purposes. They wanted to use a strategy known as multiple reaction monitoring (MRM), which collects MS data during certain time points associated with target molecules.
This approach ensures that scientists collect the data they need without being bottlenecked by extraneous spectra. But small changes to the chromatography system can lead to big changes in retention times. Those changes can, in turn, lead to the MS equipment not collecting spectra from the correct compounds. Fixing this problem requires manual adjustments that are time-consuming and expensive.
To solve that issue, Lemoine and Vandenesch turned to an approach called scout-triggered MRM (stMRM), which relies on specific compounds called marker compounds to trigger the MS to begin capturing data without needing to calculate the retention time in advance.
Scientists spike a sample with marker peptides (or other biomolecules) to serve as marker compounds to trigger the acquisition of mass spectra of the chemicals of interest. Sophisticated software algorithms are programmed to trigger the data acquisition.
The heavy reliance of the stMRM method on computer-based analytics means that scientists can change the way they control these instruments digitally. In addition, by using the same computer code, two independent laboratories can conduct the same analysis and get the same results.
Code for the future
Programming expertise at Sciex prompted a close collaboration in 2015 with Lemoine and Vandenesch to develop stMRM.
The system was tested using a biological sample that contained over 200 peptides and 10 to 12 marker compounds. The Sciex researchers were able to detect 90pc of the compounds that were expected to be in the original sample. The missed compounds were not the fault of the equipment or software, but rather sample degradation during shipping.
By optimising the liquid chromatography step, the scientists eventually were able to prepare a biological sample in less than 15 minutes and perform an analysis in less than 12 minutes. In fact, stMRM shows so much promise that the French team won a major grant to develop a system that would help precision medicine.
Although much of the work in antimicrobial resistance detection and stMRM has focused on peptides, there is the possibility of extending this technology to other biomolecules. And since so much of the stMRM process can potentially be automated, researchers should be able to make the system easier to use for a wider range of compounds.
Many of these advances will depend on improving the code to increase capability while simultaneously considering all the different biochemical variables that can affect the ability of stMRM to separate the signal from the noise.
This method in a research setting can provide deeper and faster information than standard PCR-based genotyping. One assay of biological samples can identify the causative bacterium as well as important epidemiological information, such as antibiotic resistance, in just 60 to 80 minutes. Another stMRM approach can analyse over 300 peptides in biological samples of sepsis-causing bacteria in just eight minutes.
With millions of people dying each year from AMR and the problem only projected to increase, there is a real need for technologies such as this new stMRM research method. Once they are sufficiently proven and refined for transition into a clinical setting with clinical-grade equipment and reagents, they have real potential to slow the development of antimicrobial resistance.
Better, faster diagnostics will enable better treatments for infections and more effective antimicrobial stewardship.
By Chris Lock
Chris Lock is vice-president of R&D at Sciex, which provides analytical technologies for the life sciences industry. Sciex is designing a user-friendly interface for clinicians and laboratory technicians and its Sciex OS software provides a platform for stMRM.
10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republic’s digest of essential sci-tech news.