Serotonylation of vascular proteins important to contraction

Abstract

Background: Serotonin (5-hydroxytryptamine, 5-HT) was named for its source (sero-) and ability to modify smooth muscle tone (tonin). The biological effects of 5-HT are believed to be carried out by stimulation of serotonin receptors at the plasma membrane. Serotonin has recently been shown to be synthesized in vascular smooth muscle and taken up from external sources, placing 5-HT inside the cell. The enzyme transglutaminase uses primary amines such as 5-HT to covalently modify proteins on glutamine residues. We tested the hypothesis that 5-HT is a substrate for transglutaminase in arterial vascular smooth muscle, with protein serotonylation having physiological function.

Methodology/principal findings: The model was the rat aorta and cultured aortic smooth muscle cells. Western analysis demonstrated that transglutaminase II was present in vascular tissue, and transglutaminase activity was observed as a cystamine-inhibitable incorporation of the free amine pentylamine-biotin into arterial proteins. Serotonin-biotin was incorporated into alpha-actin, beta-actin, gamma-actin, myosin heavy chain and filamin A as shown through tandem mass spectrometry. Using antibodies directed against biotin or 5-HT, immunoprecipitation and immunocytochemistry confirmed serotonylation of smooth muscle alpha-actin. Importantly, the alpha-actin-dependent process of arterial isometric contraction to 5-HT was reduced by cystamine.

Conclusions: 5-HT covalently modifies proteins integral to contractility and the cytoskeleton. These findings suggest new mechanisms of action for 5-HT in vascular smooth muscle and consideration for intracellular effects of primary amines.

Figure 1. TG protein and activity are observed in rat aorta.

A. Immunohistochemical localization of TG II and the classical protein product containing an N(ε)-(γ-glutamy)-lysine in band in normal rat thoracic aorta. Arrows point to positive staining that is present in samples incubated with primary antibody (primary) but is lost when the primary antibody is removed from the reaction (no primary). Representative of four (4) separate animals. L = lumen. B. Western analysis demonstrating the presence of TGII in homogenate of the rat aorta (arrow). Each lane represents a different animal. Positive control is the rat liver. C. TG activity assay in homogenate from rat thoracic aorta. Samples were incubated in normal TG buffer (in the absence and presence of the TG substrate BAP (−BAP, +BAP respectively), or TG inhibitor cystamine (10 mM). Representative of N>18.

Figure 2. Endogenous 5-HT and TGII activity exists in rat aorta.

A. Left: HPLC measurement of endogenous 5-HT and 5-HIAA in rat thoracic aorta for the number of animals indicated in parentheses. Bars and vertical lines represent means±SEM. Right: Recognition of proteins by an anti-5-HT antibody in normal rat aortic homogenate not exposed to the monoamine oxidase inhibitor pargyline or exogenous 5-HT. Each lane represents a separate animal, and this represents an N = 8. B. Left: TG activity assay using 5-HT-biotin as substrate in homogenates of rat thoracic aorta. Samples were incubated with or without the TG inhibitor cystamine (10 mM), as well as in the presence of cystamine in zero calcium TG buffer. Representative of N>30 different samples. Right: Lack of biotinylated proteins when biotin alone was used as a substrate in the absence or presence of cystamine. Each lane represents a different animal, representative of N = 8.

Figure 3. Protein serotonylation is time- and concentration-dependent in aortic homogenates and can be competed against by excess 5-HT.

A. Time course of protein modification for 5-HT-biotin and BAP as substrates. Representative of N = 3 separate experiments. B. Concentration-response curve for aortic protein serotonylation. Representative of N = 5 separate experiments. C. Left: Ability of weight/weight excess of 5-HT to compete off 5-HT-biotin (12. 7 µM) in a TG reaction. Lower blot demonstrates that protein was loaded equally into all lanes as observed through equal α-actin expression. Right panel shows densitometry for the band at 40–42 kDa protein. Representative of N = 6 separate experiments.

Figure 4. Aortic smooth muscle cells take up 5-HT and 5-HT-biotin; 5-HT-biotin is incorporated into cellular proteins.

A. 5-HT uptake of rat aortic smooth muscle cells in the presence of the serotonin transporter inhibitor fluoxetine (1 µM). 5-HIAA levels were nearly zero, and are thus not visible but are marked on the graph. Data are from N = 6 separate aortic explants. Bars and vertical lines represent means±SEM. B. Serotonylation of cytosolic proteins upon incubation of 5-HT-biotin with rat aortic vascular smooth muscle cells in the presence and absence of the serotonin transporter inhibitor fluoxetine (1 µM); lower blot is α-actin loading control. Representative of cells from 6 different aortic explants.

Figure 5. α-actin is serotonylated in aortic smooth muscle cells and inhibition of TG activity reduces aortic contraction to 5-HT.

A. Immunoprecipitation of smooth muscle α-actin from rat aortic homogenates exposed to 5-HT-biotin in a standard transglutaminase reaction. Blots were developed using a streptavidin secondary (top), or exposed to a primary antibody against α-actin (bottom) and developed using standard horseradish peroxidase secondary antibody. Representative of N = 6 different experiments. B. Effect of vehicle (filled symbol) and cystamine (0.1–1 mM; open symbol) on 5-HT (top) and KCl (bottom)-induced contraction in isolated rat aorta. * indicates statistical difference from vehicle-incubated values. Points and vertical lines represent means±SEM for number of animals in parentheses.

Figure 6. 5-HT and 5-HT-biotin localize to α-actin and are incorporated into proteins.

A. Immunocytochemistry of aortic smooth muscle cells incubated with exogenous 5-HT (12.7 µM; left) or 5-HT biotin (12.7 µM; right) and α-actin for 1 hour prior to fixation and visualization using an antirabbit fluorescent secondary (for 5-HT) or streptavidin-conjugated secondary (for 5-HT biotin). Representative of four different aortic explants. B. Effect of cystamine (10 mM) on 5-HT-biotin localization in aortic smooth muscle cells. Representative of four different aortic explants.

Figure 7. Serotonylation occurs in multiple tissues, smooth muscular and non-smooth muscular.

Serotonylation of proteins from homogenates of rat thoracic aorta, rat stomach fundus, rat small intestine and rat cerebral cortex. The right half of the blot shows inhibition of serotonylation by the TG inhibitor cystamine. Representative of three separate experiments, each using 2 samples from different animals.