Mass spectrometry imaging of triglycerides in biological tissues by laser desorption ionization from silicon nanopost arrays

Abstract Mass spectrometry imaging (MSI) is used increasingly to simultaneously detect a broad range of biomolecules while mapping their spatial distributions within biological tissue sections. Matrix‐assisted laser desorption ionization (MALDI) is recognized as the method‐of‐choice for MSI applications due in part to its broad molecular coverage. In spite of the remarkable advantages offered by MALDI, imaging of neutral lipids, such as triglycerides (TGs), from tissue has remained a significant challenge due to ion suppression of TGs by phospholipids, e.g. phosphatidylcholines (PCs). To help overcome this limitation, silicon nanopost array (NAPA) substrates were introduced to selectively ionize TGs from biological tissue sections. This matrix‐free laser desorption ionization (LDI) platform was previously shown to provide enhanced ionization of certain lipid classes, such as hexosylceramides (HexCers) and phosphatidylethanolamines (PEs) from mouse brain tissue. In this work, we present NAPA as an MSI platform offering enhanced ionization efficiency for TGs from biological tissues relative to MALDI, allowing it to serve as a complement to MALDI‐MSI. Analysis of a standard lipid mixture containing PC(18:1/18:1) and TG(16:0/16:0/16:0) by LDI from NAPA provided an ~49 and ~227‐fold higher signal for TG(16:0/16:0/16:0) relative to MALDI, when analyzed without and with the addition of a sodium acetate, respectively. In contrast, MALDI provided an ~757 and ~295‐fold higher signal for PC(18:1/18:1) compared with NAPA, without and with additional Na + . Averaged signal intensities for TGs from MSI of mouse lung and human skin tissues exhibited an ~105 and ~49‐fold increase, respectively, with LDI from NAPA compared with MALDI. With respect to PCs, MALDI provided an ~2 and ~19‐fold increase in signal intensity for mouse lung and human skin tissues, respectively, when compared with NAPA. The complementary coverage obtained by the two platforms demonstrates the utility of using both techniques to maximize the information obtained from lipid MS or MSI experiments.