z-logo
Premium
Cluster Analysis Identifies Phosphatidylcholine Species in the Liver and Heart as Prediction Hubs in Ossabaw Swine with Cardio‐Metabolic Heart Failure
Author(s) -
Li Jinpu,
Rau Cristoph,
Olver T. Dylan,
Franke Jacob,
Crockett Emery,
Ivey Jan,
Thorne Pamela,
Domeier Timothy,
Padilla Jaume,
Rector R. Scott,
Wang Yibin,
Ford David,
Emter Craig,
Leary Emily
Publication year - 2021
Publication title -
the faseb journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.709
H-Index - 277
eISSN - 1530-6860
pISSN - 0892-6638
DOI - 10.1096/fasebj.2021.35.s1.04226
Subject(s) - heart failure , medicine , cluster (spacecraft) , cardiology , phosphatidylcholine , biology , computer science , biochemistry , phospholipid , membrane , programming language
We recently developed an Ossabaw swine model (Western Diet and aortic banding; WD‐AB) that exhibits features consistent with cardio‐metabolic heart failure (HF). This pre‐clinical model has a pattern of findings possibly useful for enhancing our understanding of the interactions between metabolic syndrome and HF. The aim of this study is to use advanced quantitative methods to reveal novel relationships and candidate predictors we hypothesize may be important to cardio‐metabolic HF. A large‐scale data repository was created containing 142 hemodynamic, structural, and functional cardiovascular measures, 413 lipid species identified using shotgun lipidomics, and 28,386 mRNA‐transcripts revealed by RNA‐seq. Repository data was generated from heart, vascular, hepatic, and blood samples collected from 2 groups of intact female Ossabaw swine: 1) lean control (CON; N=5); and 2) WD‐AB (N=4). Logistic regression models with AUC scores identified a subset of measurements predictive of HF status that included 1,237 measurements (4.3%). Log2 fold changes were then calculated and statistically significant changes (p<0.05) with absolute value log2 fold changes >1 were considered. Pearson correlations for these 153 measurements were analyzed to identify redundant information using k‐medoids clustering analysis with an average silhouette approach. Twenty‐eight left ventricle (LV) genes and 125 lipids in 2 clusters (Fig. 1 Circos plot) with center points were identified. Cluster 1 (7 mRNA‐transcripts, 70 lipids) was composed of strong correlations with center point phosphatidylcholine (PC) alkyl ether (a) species 32 from the liver (LIVER_PC­_32.0a_754) that increased 10‐fold in the WD‐AB group (P<0.05; 25±4 CON vs. 277±56 WD‐AB pmols/mg). Cluster 2 (21 mRNA‐transcripts, 55 lipids) had center point PC species 38.4a from the LV (HEART_PC_38.4a_830) that increased ~3‐fold in the WD‐AB group (P<0.05; 37±3 CON vs. 102±3 WD‐AB pmols/mg). Identified PCa species containing arachidonic acid are storage depots for bioactive oxylipids (eicosanoids) and a precursor for platelet activating factor production. Increased alkyl ether phospholipids may also indicate increased peroxisome proliferation (the site of alkyl ether biosynthesis) and the potential to increase unknown peroxisome proliferator‐activated receptor (PPAR) endogenous ligands. In conclusion, PCa species identified from the heart and liver of WD‐AB animals may be underappreciated and have predictive significance for cardio‐metabolic HF. Ongoing work includes enrichment analysis to determine biological processes underlying clusters.

This content is not available in your region!

Continue researching here.

Having issues? You can contact us here