Evaluation of Formalin-Fixed and FFPE Tissues for Spatially Resolved Metabolomics and Drug Distribution Studies

Abstract

Fixation of samples is broadly used prior to the histological evaluation of tissue samples. Though recent reports demonstrated the ability to use fixed tissues for mass spectrometry imaging (MSI) based proteomics, glycomics and tumor classification studies, to date comprehensive evaluation of fixation-related effects for spatially resolved metabolomics and drug disposition studies is still missing. In this study we used matrix assisted laser desorption/ionization (MALDI) and desorption electrospray ionization (DESI) MSI to investigate the effect of formalin-fixation and formalin-fixation combined with paraffin embedding on the detectable metabolome including xenobiotics. Formalin fixation was found to cause significant washout of polar molecular species, including inorganic salts, amino acids, organic acids and carnitine species, oxidation of endogenous lipids and formation of reaction products between lipids and fixative ingredients. The slow fixation kinetics under ambient conditions resulted in increased lipid hydrolysis in the tissue core, correlating with the time-dependent progression of the fixation. Paraffin embedding resulted in subsequent partial removal of structural lipids resulting in the distortion of the elucidated biodistributions.

Keywords: DESI; FFPE; MALDI; mass spectrometry imaging; sample preparation.

Figure 1

Representative mean spectra of (a) fresh-frozen, (b) formalin-fixed and (c) formalin-fixed, paraffin embedded rat liver specimens analyzed by DESI-MSI in positive and negative ion mode, respectively. All tissues were collected from the same animal.

Figure 2

Mean-Fold changes of metabolite abundances as determined by DESI-MSI of fresh-frozen and formalin-fixed rat kidneys and fresh-frozen, formalin-fixed and formalin-fixed, paraffin embedded rat liver specimens respectively. Mean-fold changes were calculated to the average of all analyzed specimens in all treatment groups of one organ. The assigned numbers in the header refer to the individual animal IDs (i.e., #1 = Animal 1). Detailed information for the annotations can be found in Table S1. Detailed information regarding statistical significance for the changes can be found in Table S2.

Figure 3

Comparison of representative fresh-frozen and formalin-fixed rat liver specimens. (a) Pixel-wise PCA of the tissue sections analyzed by DESI-MSI in negative ion mode. PC1 (red) contributed 16.65% PC2 (green) 9.38% and PC3 (blue) 5.52% of the total observed variance. (b) The regional differences found within the PC1–3 were found to correlate with the degradation of phospholipids in the tissue core and occurrence of the hydrolysis products lyso-lipids (18:0) and FA(20:4). (c) LDH stain of a frozen liver section compared to (d) LDH stain of a tissue section after fixation of the tissue for 24 h. (e) LDH stained tissue sections of rat liver punches in various sizes after 1 h of fixation.

Figure 4

Effects of the sample collection and preparation on (a) representative distributions of the drugs and (b) mean abundance within the tissue sections for all biological replicates as determined by DESI-MSI. Statistical significance one-way ANOVA for liver samples, two-tailed t-test for kidneys. ns = p (>0.05), * = p (<0.05).

Figure 5

(a) Hepatic lobule of a fresh frozen liver section outlined by PC(36:2) (b) Terfenadine (green) in relative distribution to Heme B as determined by high resolution MALDI-MSI. (c) Distribution of Fexofenadine (active metabolite) compared to Heme B (d) RGB overlay of Terfenadine, Fexofenadine and Heme B. The box highlights the pixel used for the profile plot. (e) Profile plot of the normalized mean abundances of terfenadine, fexofenadine compared to PC(36:2) outlining the hepatic lobule and Heme B marking the central vein. Data is represented as mean of 3 separate lines, normalized to the highest pixel and cubically smoothed over 3 neighboring pixels to compensate for pixel-to-pixel variability.

Figure 6

(a) Current sample collection workflow including separation of samples during collection. (b) facilitated workflow with snap freezing of all specimens during sample collection and subsequent fixation of specimens for histological evaluation.