Proteomics in gastroparesis: unique and overlapping protein signatures in diabetic and idiopathic gastroparesis

Madhusudan Grover, Surendra Dasari, Cheryl E Bernard, Lakshmikanth L Chikkamenahalli, Katherine P Yates, Pankaj J Pasricha, Irene Sarosiek, Richard McCallum, Kenneth L Koch, Thomas L Abell, Braden Kuo, Robert J Shulman, Simon J Gibbons, Travis J McKenzie, Todd A Kellogg, Michael L Kendrick, James Tonascia, Frank A Hamilton, Henry P Parkman, Gianrico Farrugia, Madhusudan Grover, Surendra Dasari, Cheryl E Bernard, Lakshmikanth L Chikkamenahalli, Katherine P Yates, Pankaj J Pasricha, Irene Sarosiek, Richard McCallum, Kenneth L Koch, Thomas L Abell, Braden Kuo, Robert J Shulman, Simon J Gibbons, Travis J McKenzie, Todd A Kellogg, Michael L Kendrick, James Tonascia, Frank A Hamilton, Henry P Parkman, Gianrico Farrugia

Abstract

Macrophage-based immune dysregulation plays a critical role in development of delayed gastric emptying in diabetic mice. Loss of anti-inflammatory macrophages and increased expression of genes associated with pro-inflammatory macrophages has been reported in full-thickness gastric biopsies from gastroparesis patients. We aimed to determine broader protein expression (proteomics) and protein-based signaling pathways in gastric biopsies of diabetic (DG) and idiopathic gastroparesis (IG) patients. Additionally, we determined correlations between protein expressions, gastric emptying, and symptoms. Full-thickness gastric antrum biopsies were obtained from nine DG patients, seven IG patients, and five nondiabetic controls. Aptamer-based SomaLogic tissue scan that quantitatively identifies 1,305 human proteins was used. Protein fold changes were computed, and differential expressions were calculated using Limma. Ingenuity pathway analysis and correlations were carried out. Multiple-testing corrected P < 0.05 was considered statistically significant. Seventy-three proteins were differentially expressed in DG, 132 proteins were differentially expressed in IG, and 40 proteins were common to DG and IG. In both DG and IG, "Role of Macrophages, Fibroblasts and Endothelial Cells" was the most statistically significant altered pathway [DG false discovery rate (FDR) = 7.9 × 10-9; IG FDR = 6.3 × 10-12]. In DG, properdin expression correlated with GCSI bloating (r = -0.99, FDR = 0.02) and expressions of prostaglandin G/H synthase 2, protein kinase C-ζ type, and complement C2 correlated with 4 h gastric retention (r = -0.97, FDR = 0.03 for all). No correlations were found between proteins and symptoms or gastric emptying in IG. Protein expression changes suggest a central role of macrophage-driven immune dysregulation in gastroparesis, specifically, complement activation in diabetic gastroparesis.NEW & NOTEWORTHY This study uses SOMAscan, a novel proteomics assay for determination of altered proteins and associated molecular pathways in human gastroparesis. Seventy-three proteins were changed in diabetic gastroparesis, 132 in idiopathic gastroparesis compared with controls. Forty proteins were common in both. Macrophage-based immune dysregulation pathway was most significantly affected in both diabetic and idiopathic gastroparesis. Proteins involved in the complement and prostaglandin synthesis pathway were associated with symptoms and gastric emptying delay in diabetic gastroparesis.

Keywords: diabetes; immune cells; inflammation; macrophages.

Conflict of interest statement

No conflicts of interest, financial or otherwise, are declared by the authors.

Figures

Graphical abstract
Graphical abstract
Fig. 1.
Fig. 1.
Heat map showing differentially expressed proteins (false discovery rate A: abundance of 73 proteins was changed in diabetic gastroparesis [n = 9 diabetic gastroparesis, 5 controls (Ctrl)]. B: abundance of 132 proteins was changed in idiopathic gastroparesis (n = 7 idiopathic gastroparesis, 5 controls). Ctrl, control; DG, diabetic gastroparesis.
Fig. 1.
Fig. 1.
Heat map showing differentially expressed proteins (false discovery rate A: abundance of 73 proteins was changed in diabetic gastroparesis [n = 9 diabetic gastroparesis, 5 controls (Ctrl)]. B: abundance of 132 proteins was changed in idiopathic gastroparesis (n = 7 idiopathic gastroparesis, 5 controls). Ctrl, control; DG, diabetic gastroparesis.
Fig. 2.
Fig. 2.
A: volcano plot showing expressed proteins. B: diabetic gastroparesis vs. controls. Idiopathic gastroparesis vs. controls. Proteins in red indicate false discovery rate (FDR) < 0.01.
Fig. 3.
Fig. 3.
Venn diagram showing unique and common differentially abundant proteins in diabetic and idiopathic gastroparesis. Thirty-three proteins were uniquely changed (17 overexpressed and 16 underexpressed) in diabetic gastroparesis as compared with controls. Ninety-two proteins were uniquely changed in idiopathic gastroparesis (66 overexpressed and 26 underexpressed). Forty proteins were common to both diabetic and idiopathic gastroparesis (15 overexpressed and 25 underexpressed in both). Protein names and log fold changes shown using colored codes.
Fig. 4.
Fig. 4.
Correlations between protein expression (normalized and log transformed), symptoms, and gastric emptying in diabetic gastroparesis. A: properdin (CFP) abundance negatively correlated with gastroparesis cardinal symptom index (GCSI) bloating. No correlations were seen between other subscores (nausea, fullness) or overall GCSI. BD: negative correlation between 4 h of gastric retention and expressions of prostaglandin G/H synthase 2 (PTGS2), complement 2 (C2), and protein kinase Cζ (PRKCZ). No correlations were seen between protein abundance, symptoms, and gastric retention in idiopathic gastroparesis. A complete list of all protein expressions in patients with diabetic and idiopathic gastroparesis is available in GEO database (GSE130672; https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE130672; Token: grwpcgikvfmvzgl). AU, arbitrary units. FDR, false discovery rate.

Source: PubMed

3
Sottoscrivi