Multi-OMICS at the single cell level using analytical scale chromatography with a multi-reflecting ToF MS

The development of cutting-edge mass spectrometers with high levels of sensitivity and resolution has led to single-cell OMICs gaining significant momentum in recent years. Now, previously unanswerable questions can be probed through the distinct measurements of single cells in lieu of a bulk cell average. However, there remains infinite scope for optimisation in single cell OMIC techniques, from sample collection to data analysis. Here we investigate, optimise and present a single cell lipidomic workflow, established with cultured human cell lines using liquid chromatography (LC) coupled to a multi-reflecting time-of-flight mass spectrometer (MS).
Several parameters were assessed and established using several different cell lines (e.g. HT-29, Caco-2 and T47-D) which were cultivated in-house. These were isolated using either fluorescence-activated cell sorting (FACS) or the single-cell picking platform (IotaSciences). Lipids were extracted using isopropanol (IPA), whilst polar metabolites were extracted using a peptide digestion protocol, providing peptides and polar metabolites. Lipids and polar metabolites were chromatographically separated using UHPLC, configured with a 2.1x100 mm CSH phenyl-hexyl reversed-phase (RP) column. The advantage of this column chemistry is that it eliminates the use of IPA whilst allowing for lipids and polar metabolites to be separated using the same column and mobile phases. The LC was coupled to a multi-reflecting time-of-flight (ToF) mass spectrometer (MS) and data collected using a data independent acquisition (DIA) schema. The subsequent data were processed using LipoStar and MARS software packages for compound identification.
Based on the LC method of 6.5 min (injection-injection), approx. 200 lipid identifications and >100 polar metabolites for a single cell. 

Identifications were based on mass accuracies of <500 ppb (precursor and fragment ions). The data revealed various lipid classes being dysregulated between cell types, for example, HT29 and Caco-2 indicated triglycerides were at elevated levels in Caco-2. Biological interpretation indicates that the free fatty acids (FFA’s) are consumed by TG’s for resynthesis. HT-29 on the other hand, had higher levels of phosphocholines (PC’s) due to the secretion of mucus from HT-29. PC’s have previously been shown to be a major lipid component of mucus.