Lipidomics Reveals a Tissue-Specific Fingerprint

Abstract

In biological systems lipids generate membranes and have a key role in cell signaling and energy storage. Therefore, there is a wide diversity of molecular lipid expressed at the compositional level in cell membranes and organelles, as well as in tissues, whose lipid distribution remains unclear. Here, we report a mass spectrometry study of lipid abundance across 7 rat tissues, detecting and quantifying 652 lipid molecular species from the glycerolipid, glycerophospholipid, fatty acyl, sphingolipid, sterol lipid and prenol lipid categories. Our results demonstrate that every tissue analyzed presents a specific lipid distribution and concentration. Thus, glycerophospholipids are the most abundant tissue lipid, they share a similar tissue distribution but differ in particular lipid species between tissues. Sphingolipids are more concentrated in the renal cortex and sterol lipids can be found mainly in both liver and kidney. Both types of white adipose tissue, visceral and subcutaneous, are rich in glycerolipids but differing the amount. Acylcarnitines are mainly in the skeletal muscle, gluteus and soleus, while heart presents higher levels of ubiquinone than other tissues. The present study demonstrates the existence of a rat tissue-specific fingerprint.

Keywords: adult rat tissues; cholesterol; glycerolipids; glycerophospholipids; lipid distribution; lipidomics; plasmalogens; sphingolipids.

Figure 1

Multivariate statistics reveal a specific lipidomic profile for each tissue. (A,B) Principal component analysis (PCA) representation of the lipidome of all the tissues in positive and negative ionization polarities. (C,D) PCA representation of the lipidomic profiles of all the tissues except for both type of adipose tissues in positive and negative ionization polarities. X: Principal component 1 (PC1), Y: Principal component 2 (PC2), Z: Principal component 3 (PC3). (E) Hierarchical clustering algorithm represented by a heat map of the lipid species detected in both polarities of ionization in all the tissues.

Figure 2

Lipid species concentration and distribution across the tissues analyzed by a targeted lipidomic approach. (A) Graph representation of average concentration in nmol/g of tissue of lipid species detected by targeted lipidomics grouped by families. Error bars denote SEM across 7–9 animals. Adipose tissue is comprised of visceral and subcutaneous tissue while gluteus and soleus comprised skeletal muscle. Kidney is represented only by the renal cortex. Statistical analysis performed was one-way ANOVA and post-hoc Tukey multiple test. **p < 0.01, and ***p < 0.001. (B) Heat map of abundance of the detected lipid species across samples. Each line represents one compound colored by its normalized abundance to internal standard, baseline and mean across samples. The scale from 2 to −2 represents the normalized abundance of each lipid in arbitrary units. *1 gangliosides and sulfatides; *2 hydroxycholesterol, desmosterol and cholesteryl esters.

Figure 3

Glycerophospholipids within mammalian tissues detected by targeted lipidomic analysis. (A) Concentration of PC, PE, PS, PI, and PG subclasses. Values are expressed as mean ± SEM from 8 to 10 animals. Statistical analysis was one-way ANOVA and post-hoc Tukey significance is represented in the bar chart, meaning a significantly different respect to SAT, b respect to VAT, c respect to soleus, d respect to gluteus, e respect to heart and f respect to kidney, g respect to liver, τ respect to all. *p < 0.05, **p < 0.01, and ***p < 0.001. (B) Relative concentration of lipids normalized per sample to the total abundance within this lipid class to obtain molar fractions. Each solid line indicated tissue from an individual rat. Gray vertical lines separate lipids by total number of acyl chain carbons. The number of double bonds is indicated below within each group.

Figure 4

Lysoglycerophospholipids within mammalian tissues detected by targeted lipidomic analysis. (A) Concentration of LPC, LPE, LPS, and LPI subclasses. Values are expressed as mean ± SEM from 8 to 10 animals. Statistical analysis was one-way anova and post-hoc Tukey significance is represented in the bar chart, meaning a significantly different respect to SAT, b respect to VAT, c respect to soleus, d respect to gluteus, e respect to heart and f respect to kidney, g respect to liver, τ respect to all. *p < 0.05, **p < 0.01, and ***p < 0.001. (B) Relative concentration of lipids normalized per sample to the total abundance within this lipid class to obtain molar fractions. Each solid line indicated tissue from an individual rat. Gray vertical lines separate lipids by total number of acyl chain carbons. The number of double bonds is indicated below within each group.

Figure 5

Ether lipid species within mammalian tissues detected by targeted lipidomic analysis. (A,C) Concentration of alkyl and alkenyl, respectively. Values are expressed as mean ± SEM from 8 to 10 animals. Statistical analysis was one-way anova and post-hoc Tukey significance is represented in the bar chart, meaning a significantly different respect to SAT, b respect to VAT, c respect to soleus, d respect to gluteus, e respect to heart and f respect to kidney, g respect to liver, τ respect to all. *p < 0.05, **p < 0.01, and ***p < 0.001. (B,D) Relative concentration of lipids normalized per sample to the total abundance within this lipid class to obtain molar fractions. Each solid line indicated tissue from an individual rat. Gray vertical lines separate lipids by total number of acyl chain carbons. The number of double bonds is indicated below within each group.

Figure 6

Structural sphingolipids and sterol lipids within mammalian tissues detected by targeted lipidomic analysis. (A,C) Concentration of SP and ST subclasses. Values are expressed as mean ± SEM from 8 to 10 animals. Statistical analysis was one-way anova and post-hoc Tukey significance is represented in the bar chart, meaning a significantly different respect to SAT, b respect to VAT, c respect to soleus, d respect to gluteus, e respect to heart and f respect to kidney, g respect to liver, τ respect to all. *p < 0.05, **p < 0.01, and ***p < 0.001. (B) Relative concentration of lipids normalized per sample to the total abundance within this lipid class to obtain molar fractions. Each solid line indicated tissue from an individual rat. Gray vertical lines separate lipids by total number of acyl chain carbons. The number of double bonds is indicated below within each group.

Figure 7

Second messenger lipid species within mammalian tissues detected by targeted lipidomic analysis. (A) Concentration of DAG, Cer, Cer1P, dhCer, Sph, and Sp1P. Values are expressed as mean ± SEM from 8 to 10 animals. Statistical analysis was one-way anova and post-hoc Tukey significance is represented in the bar chart, meaning a significantly different respect to SAT, b respect to VAT, c respect to soleus, d respect to gluteus, e respect to heart and f respect to kidney, g respect to liver, τ respect to all. *p < 0.05, **p < 0.01, and ***p < 0.001. (B) Relative concentration of lipids normalized per sample to the total abundance within this lipid class to obtain molar fractions. Each solid line indicated tissue from an individual rat. Gray vertical lines separate lipids by total number of acyl chain carbons. The number of double bonds is indicated below within each group.

Figure 8

Bioenergetic lipid species within mammalian tissues detected by targeted lipidomic analysis. (A) Concentration of FAC, TAG, CE, and oxCE. Values are expressed as mean ± SEM from 8 to 10 animals. Statistical analysis was one-way anova and post-hoc Tukey significance is represented in the bar chart, meaning a significantly different respect to SAT, b respect to VAT, c respect to soleus, d respect to gluteus, e respect to heart and f respect to kidney, g respect to liver, τ respect to all. *p < 0.05, **p < 0.01, and ***p < 0.001. (B) Relative concentration of lipids normalized per sample to the total abundance within this lipid class to obtain molar fractions. Each solid line indicated tissue from an individual rat. Gray vertical lines separate lipids by total number of acyl chain carbons. The number of double bonds is indicated below within each group.

Figure 9

Cluster analysis of lipid profiles. Left: heat map of Pearson correlation matrix across the 652 targeted lipids with corresponding hierarchical tree. White boxes indicate lipid clusters chosen for further analysis. Right: Concentration of lipid clusters across all the tissues analyzed. For the information about the lipid species in each lipid cluster see Table S3.