Using a search-based “bootstrapping” method, researchers at the National Institute of Standards and Technology (NIST) have more than doubled a reference library that identifies compounds in milk.
The first version of the human milk oligosaccharide (HMO) mass spectral library, released in 2018, comprised 74 oligosaccharides, a specific type of carbohydrate found in milk. The 2020 version identifies an additional 80 HMOs from human milk, and extends coverage to a variety of mammals, reporting yet another 65 oligosaccharides.
Oligosaccharides are known to have a biological effect, such as providing energy for growing babies or contributing to organ development, but researchers have yet to understand exactly how that occurs. If scientists can determine what oligosaccharides are present in milk, they can begin to understand how the carbohydrate affects cells, tissues and other biological processes.
“The goal is to identify any oligosaccharide one can find in milk,” Stephen Stein, who has led the development of the NIST Spectral Library for more than 20 years, told Laboratory Equipment. “If you don’t know what the molecules are, you can’t really say much about them. We want to identify everything in there, everything we can find with modern technology and mass spectrometry.”
With nearly 1.5 million spectra, the NIST Mass Spectral Library is among the largest commercially available libraries, used by scientists in a variety of industries. In 2018, Stein pioneered a hybrid search algorithm that enables unknown similar compounds to be found even when there isn’t an exact match in the library. It was this search tool that ended up being the catalyst for the expansion of the HMO library, said Connie Remoroza, a NIST chemist.
“I built the HMO library in the first place using the hybrid search of oligosaccharide spectra in the NIST17 library. For the second part, I searched unknowns in milk against the HMO library and the unknown library through hybrid search. We found we could expand a lot on oligosaccharides based on searching from these two databases,” Remoroza explained to Laboratory Equipment.
Using NIST Mass Spectral Libraries and LCMS, Remoroza and her team identified an additional 80 oligosaccharides, doubling the initial library. They also found new compounds that had previously never been reported in milk, including one type of oligosaccharide that contains 15 monosaccharide units, the building blocks of carbohydrates.
And that’s just in human milk. In mammal milk, the researchers identified another 65 oligosaccharides. It was an incidental conversation with a co-worker that alerted Remoroza to the fact that there was a treasure trove of mammal milk just 18 miles down the road from her NIST office.
The Smithsonian's National Zoo and Conservation Biology Institute provided milk samples from a Saanen goat, Asian water buffalo and African lion, while NIST had in-house milk samples from cows. While prior research had been done on these animal samples, relatively few oligosaccharides had been reported in the literature.
“Many oligosaccharides are now known because of improved sensitivity of the mass spectrometers, combined with NIST search software," Remoroza said.
Unsurprisingly, the HMO library is of special interest to infant formula manufacturers. But, Stein says, it has implications across a wide variety of research interests, from biological to food.
“In the food industry, once they know this is possible, it will open up doors for them to do their analysis in ways we don’t even know about,” he said. “It’s like a microscope, allowing you to see more things than you have seen before and allowing other ways to solve problems.”
Remoroza and her team will continue to identify the different types of oligosaccharides in human and nonhuman mammalian milk, including black and white pigs, rhesus monkeys and dolphins. In addition to expansion of the library, the scientists are also interested in analyzing glycopeptides in milk, which is very difficult as they hide the mass of molecules, rendering traditional mass spectrometry methods ineffective.
“Identifying all of the glycans, looking at reproducibility and then putting them in a library is very challenging,” Stein said, “but, we have new methods for the identification of glycans. Once we get them into a library, people will then be able to routinely and simply identify these entities.”
The larger NIST Mass Spectral Library was just updated last month to add 6,000 human metabolites, 8,000 plant metabolites, 2,000 drugs, 1,000 pesticides and 1,000 lipids.
Photo: Remoroza working with a mass spectrometer in the NIST laboratory. Credit: Rich Press/NIST