Purchase this article with an account.
Ankita Kotnala, David M G Anderson, Nathan Heath Patterson, Jeffery D Messinger, Christine Curcio, Kevin L Schey; Sample preparation effects on retina lipid analysis by MALDI imaging and LC-MS technologies.. Invest. Ophthalmol. Vis. Sci. 2019;60(9):2986. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
In age-related macular degeneration (AMD), ample evidence supports a role of lipids and associated pathways in deposit biogenesis and degeneration of outer retinal cells. Imaging Mass Spectrometry (IMS) localizes lipids in tissues based on molecular weights, in a label-free manner. Liquid Chromatography Tandem Mass Spectrometry (LC-MS/MS) confirms the molecular identity of lipids observed with IMS. Paraformaldehyde fixation improves morphology of delicate retina but may impact signal detection. We compared data from fixed and freshly frozen prepared tissues using MALDI IMS and from two lipid extraction procedures for LC-MS analysis in positive and negative ion mode to establish if fixation protocols adversely affect analyte detection or introduce artefactual signals.
Human and bovine eye pairs were prepared via two methods. OS was dissected, frozen on liquid nitrogen vapor, embedded in carboxymethyl cellulose (CMC) and stored at -80 degree celsius. OD was fixed in for 24 hours each in 4% then 1% paraformaldehyde at 4 degree celsius before processing as with OS. Tissue cryo-sections were analyzed via MALDI IMS. For LC-MS/MS analysis, embedded tissues were subjected to Folch extraction (two-phase) and an MMC (MeOH/MTBE/CHCl3) procedure (one-phase). Untargeted lipidomics data were analyzed by XCMS, MS-DIAL and Lipid Match. MALDI and LC-MS/MS data were subjected to statistical analysis.
In the same donor retina lipid signals observed by MALDI IMS and LC-MS were differentially abundant in fresh-frozen vs fixed tissues. Partial least square analysis, a multivariate approach, of MALDI IMS confirmed the data are globally separable with minimal overlap between the samples in the resulting score plots. For LC-MS, 217 and 97 confirmed lipid identities were provided by MMC and the Folsch extraction procedures, respectively for freshly frozen retina. For paraformaldehyde-fixed retina, MMC-treated tissue yielded 241 confirmed lipid identities, and Folch extraction yielded 197 confirmed lipids respectively.
The tissue preparation and extraction procedure alter the molecular ion intensities in human donor retina as analysed by IMS and LC-MS/MS.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.
This PDF is available to Subscribers Only