The heparin-binding proteome in normal pancreas and murine experimental acute pancreatitis

Abstract

Acute pancreatitis (AP) is acute inflammation of the pancreas, mainly caused by gallstones and alcohol, driven by changes in communication between cells. Heparin-binding proteins (HBPs) play a central role in health and diseases. Therefore, we used heparin affinity proteomics to identify extracellular HBPs in pancreas and plasma of normal mice and in a caerulein mouse model of AP. Many new extracellular HBPs (360) were discovered in the pancreas, taking the total number of HBPs known to 786. Extracellular pancreas HBPs form highly interconnected protein-protein interaction networks in both normal pancreas (NP) and AP. Thus, HBPs represent an important set of extracellular proteins with significant regulatory potential in the pancreas. HBPs in NP are associated with biological functions such as molecular transport and cellular movement that underlie pancreatic homeostasis. However, in AP HBPs are associated with additional inflammatory processes such as acute phase response signalling, complement activation and mitochondrial dysfunction, which has a central role in the development of AP. Plasma HBPs in AP included known AP biomarkers such as serum amyloid A, as well as emerging targets such as histone H2A. Other HBPs such as alpha 2-HS glycoprotein (AHSG) and histidine-rich glycoprotein (HRG) need further investigation for potential applications in the management of AP. Pancreas HBPs are extracellular and so easily accessible and are potential drug targets in AP, whereas plasma HBPs represent potential biomarkers for AP. Thus, their identification paves the way to determine which HBPs may have potential applications in the management of AP.

Fig 1. Normal pancreas (NP) and caerulein-induced acute pancreatitis (AP).

Representative images of H&E stained histology slides of A) NP with intact pancreas architecture and B) AP showing marked oedema, inflammatory cell infiltration and acinar cell necrosis. Mean serum amylase levels in (C) NP and (D) AP in each experiment consisting of 16 individuals.

Fig 2. Preparation of a plasma membrane enriched fraction.

Coomassie-stained SDS-PAGE gel of (A) NP and (B) AP samples obtained during homogenisation and fractionation by sequential steps of centrifugation. Nu = nuclear pellet; S1 = post-nuclear supernatant; Mt = mitochondrial pellet; S2 = post-mitochondrial supernatant; C = cytosol (post-microsomal supernatant); W = wash of the microsomal pellet; Mc = microsomal pellet. Coomassie-stained SDS-PAGE and western blot analysis of 10 fractions (F1-F11) from the microsomal pellet after flotation on a sucrose gradient (0.25–2 M) in (C) NP and (D) AP. Fractions are ordered depending on their equilibrium density from light (left) to heavy (right). The enrichment of plasma membrane was assessed by western blot using an antibody against caveolin-1, which is a specific plasma membrane marker. Full Western blots in S1 Fig.

Fig 3. Heat map depicting the variation across the biological and technical replicates in extracellular pancreas HBPs.

The rows represent the various biological replicates in normal pancreas (NP) and acute pancreatitis (AP), while the columns represent proteins. Red represents over expression and green represents under expression. Biological replicate number is denoted as "BioRep" and technical replicate number as "TechRep". Hierarchical clustering was performed on both column data, to cluster the changes in protein expression, and row data, which displays the variation between samples. The stability of the instrument platform is shown in that the lowest branch of the sample variation dendrogram correctly represents the technical replicates of each sample. The next level of the dendrogram correctly separates the biological condition of the sample indicating repeatable protein expression differences between the biological conditions. A higher degree of variability is observed in the AP samples presumably reflecting the systemic effects of AP.

Fig 4. The heparin-binding putative protein interactome in normal pancreas (NP) constructed using STRING 10.5.

Nodes or HBPs are connected by protein-protein interactions known as ‘edges’.

Fig 5. The heparin-binding putative protein interactome in acute pancreatitis (AP) constructed using STRING 10.5.

Nodes or HBPs are connected by protein-protein interactions known as ‘edges’.

Fig 6. Heat map depicting the variation across the biological and technical replicates in plasma HBPs.

The rows represent the various biological replicates from plasma in health (NP) and acute pancreatitis (AP), while the columns represent proteins. Red represents over expression and green represents under expression. Biological replicate number is denoted as "BioRep" and technical replicate number as "TechRep". Hierarchical clustering was performed on both column data, to cluster the changes in protein expression, and row data, which displays the variation between samples.