PALM SPRINGS, CALIFORNIA (GenomeWeb) – Scientists from the Colorado School of Mines have developed a method to identify bacteria at the strain level using MALDI-TOF mass spectrometry to characterize fatty acids from the phospholipid cell membrane.
Kent Voorhees and Chris Cox presented on their method this week at the 2016 Mass Spectrometry: Applications to the Clinical Lab conference. Dubbed metal oxide laser ionization (MOLI) mass spec, they demonstrated an ability to distinguish bacterial species at a taxonomic level that mass spec-based protein profiling often struggles with.
"We're not trying to overthrow the protein profiling paradigm but we want to point out there's usually hurdles and potential pitfalls," Cox told GenomeWeb. MALDI-based protein profiling can do genus-level and "passable" species-level bacterial identification, he said, "but that's where the capability ends."
"If you want to do antibiotic-resistance profiling, you have to use a whole different technology," Cox added, and that method could be MOLI.
Voorhees and Cox led a study comparing their fatty acid profiling to protein profiling, published last year in Nature Scientific Reports. Analyzing bacteria from Enterobacteriaceae, Acinetobacter, and Listeria, they reported 100 percent correct classification at the species-level and 98 percent at the strain-level. Protein profiling only provided 32 percent accuracy for species-level classification in Enterobacteriaceae, 54 percent in Acinetobacter, and 67 percent in Listeria.
MOLI mass spec uses metal oxide nanoparticles to simultaneously cleave the fatty acids from the phospholipid molecules and ionize them for analysis on a MALDI-TOF instrument. "It's an in situ catalytic event as opposed to co-crystallization and ionization," Cox said. "We're actually doing some chemistry on the plate."
Several metal oxides will work, and Voorhees noted that all types of calcium oxides have worked so far. The study published last year used cerium oxide.
Though they still have to work out the specifics, Cox hypothesized that the catalytic process strips a bond in the phospholipid, freeing the fatty acid.
Voorhees said his lab stumbled upon the method looking for a compound with low molecular weight for organic matrices used in mass spec. Much of the foundational research was done to characterize biofuels through fatty acid profiles, in collaboration with the US Department of Energy's National Renewable Energy Laboratory. Cox said the very first substance used to test the technique was Crisco.
Voorhees's lab has decades of experience in mass spec-based bioprofiling, both using proteins and fatty acids, but only turned their fatty acid profiling towards bacteria in recent years.
"We were experimenting with [bacterial] protein profiling with the Bruker MALDI Biotyper system and running into issues with closely related members of various different kinds of bacteria," Cox said. "We started noticing issues where we would get things misidentified or [protein profiling] wouldn't ID them at all."
They tried out their fatty acid method, and by 2013, Voorhees and Cox had shown they could indeed identify bacteria with it, publishing a study in the Journal of Mass Spectrometry.
The method works on any MALDI instrument, which is a big potential benefit, since the US Food and Drug Administration has already approved the Bruker Biotyper and BioMérieux's Vitek MS instruments for clinical use.
"All these other MALDI instruments that have been sold for protein profiling, some can be used directly or modified to do MOLI," Voorhees said.
As with protein profiling, the identification step involves comparing the fatty acid profile to a library of reference spectra, which the researchers have been building for years.
MOLI spec joins another lipid-based bacterial identification method presented at MSACL. At last year's conference, Imperial College London researcher Nicole Strittmatter presented work on using lipid profiling to identify bacteria from human colorectal tissue samples using rapid evaporative ionization mass spec (REIMS).
But Voorhees and Cox suggested that the comparison between the methods were superficial. One, theirs uses fatty acids, as opposed to higher molecular weight intact lipids. More importantly, REIMS is a "completely different" technology, hasn't penetrated the clinical lab market as well as MALDI has, and isn't CE marked or FDA approved for clinical use.
Even though it still relies on bacterial culture, MOLI could be faster than existing culture-based methods of bacterial identification, taking about 19 hours, Voorhees said, compared to 48 or more for other methods. Of the 19 hours, culturing accounts for 18, so faster culturing could improve the method.
Cox suggested that there's still a lot of work left for him and Voorhees to do to establish MOLI mass spec as a clinically useful method. He's working with collaborators to look closely at several strains of bacteria and validate MOLI's classification ability.
"Salmonella, in comparison to Escherichia coli, is notoriously difficult," he said. "They're just so closely related." He suggested doing experiments to take a large number of Salmonella strains and test if MOLI could tell them apart. He added that his lab is already doing this with Pseudomonas and Enterococci.
Along these lines, MOLI has shown promise in separating drug-resistant Staphylococcus aureus from susceptible strains. Voorhees presented preliminary data at this week's conference suggesting that the fatty acid profiles were different and that ratios between specific fatty acids could provide a fingerprint.
"What that means biologically is the subject of ongoing research," Cox said.