Mass Spec Roundup: Lipids, Deep Learning, and Coca-Cola… 

CD1c’s Side Hustle 

An international team has uncovered an unexpected way that the immune system presents lipid antigens to T cells, revealing that the antigen-presenting molecule CD1c can display lipids not only from the top of its binding groove, but also sideways. Published in Nature Communications, the discovery challenges long-held assumptions about how antigens are positioned for immune recognition. 

“Much of immunology has been built around the idea that immune recognition follows one fixed arrangement,” said Adam Shahine in a Monash University press release. “This work shows that the immune system is more flexible than we assumed, and that there are additional ways immune cells can ‘see’ what’s around them.” 

Using high-resolution structural imaging and mass spectrometry, the team showed that CD1c can bind dozens of lipid types and, in some cases, hold multiple lipids at once. This allows part of a lipid molecule to protrude sideways rather than directly toward the T cell receptor. Rather than obstructing immune recognition, the configuration still enables effective T cell scanning. 

First author Thinh-Phat Cao said the findings reveal CD1c as “a complex molecular player with unique properties in immune recognition,” helping explain how the immune system handles chemically diverse lipid antigens. 

DeepMet Predicts Unknown Mammalian Metabolites for MS Validation 

Despite decades of effort, much of the mammalian metabolome remains uncharted. Untargeted mass spectrometry routinely detects thousands of molecular features that cannot be assigned a structure, leaving large gaps in metabolic maps. A new Nature study tackles this problem by reversing the usual workflow: instead of identifying metabolites after detection, the researchers first predict which metabolites are likely to exist. 

The team developed DeepMet, a chemical language model trained on known mammalian metabolites. By learning structural patterns and biosynthetic logic, the model generates metabolite-like molecules that are chemically plausible but absent from current databases. These candidates are then prioritized and searched for in experimental mass spectrometry data. 

Using this approach, the researchers identified and validated dozens of previously unrecognized mammalian metabolites across tissues and biofluids. The strategy allowed them to navigate what the authors describe as metabolomic “chemical ‘dark matter’,” focusing experimental efforts on structures most likely to be biologically real. 

Coupling AI-driven prediction with high-resolution mass spectrometry, the study outlines a scalable route toward more complete and interpretable metabolomic atlases. “We demonstrate that DeepMet has learned the metabolic logic embedded within the structures of known metabolites and can leverage this understanding to anticipate the existence of metabolites absent from existing metabolic maps,” the authors write.  

LC-MS/MS Quantifies Emerging Contaminants in Water 

Researchers have developed a robust LC–MS/MS workflow that enables sensitive, multi-class detection of emerging contaminants across a wide range of water matrices. Combining solid-phase extraction with targeted tandem mass spectrometry, the method delivers broad chemical coverage while maintaining the selectivity needed for ultra-trace analysis. 

The approach uses multiple-reaction monitoring (MRM) on a triple quadrupole mass spectrometer to quantify 39 contaminants – including pharmaceuticals, pesticides, and industrial chemicals – in a single run. Detection limits reached the low nanogram-per-liter to sub-nanogram-per-liter range, with strong linearity and reproducibility across bottled water, tap water, surface water, seawater, and wastewater. 

Crucially, the LC–MS/MS method was validated using real environmental samples. Wastewater influent showed the highest contaminant diversity and concentrations, while treated effluent still contained persistent compounds at measurable levels, highlighting both the sensitivity of tandem mass spectrometry and the analytical challenges facing water treatment systems. 

According to the authors, the workflow is “suitable for routine monitoring, source tracking, regulatory compliance, and risk assessment applications requiring ultra-trace sensitivity and broad chemical coverage.” 

Newsletters

Receive the latest analytical science news, personalities, education, and career development – weekly to your inbox.

I have read and understand the Privacy Notice*

Subscribe