5-Hydroxytryptamine (5HT) Receptors in the Heart Valves of Cynomolgus Monkeys and Sprague-Dawley Rats

Abstract

5-Hydroxytryptamine-2B receptor (5HT2BR) stimulation is known to cause fibroblast mitogenesis, and the mitogenic effect has been proposed to trigger valvular heart disease in humans. In this study, we used real-time polymerase chain reaction (TaqMan) to quantify transcript levels of 5HT2B, 5HT2C, and 5HT1B receptors and immunohistochemistry (IHC) to detect the tissue localization of these receptors in the normal heart valves of cynomolgus (CM) monkeys and Sprague-Dawley (S-D) rats. In both species, positive immunostaining was noted for 5HT1B and 5HT2B receptors in mitral, tricuspid, aortic, and pulmonary valves, and the cell types showing positive staining were interstitial cells and endothelial cells lining the valve leaflet. In CM monkeys, 5HT2CR was expressed only in the endothelial cells lining the leaflet, whereas S-D valves were negative for this receptor. IHC results were correlated with 5HT2B and 5HT1B receptor transcripts for all four valves. However, 5HT2C receptor transcripts were lower than 5HT2B or 5HT1B in all CM monkey valves, whereas 5HT2C transcripts were below the level of detection in any of the S-D rat valves. Our data showed the expression of 5HT2B, 5HT1B, and 5HT2C receptors in the normal heart valves of CM monkeys and S-D rats, and IHC and TaqMan techniques may be used to study the potential mechanism of compounds with 5HT2BR agonist activity.

Keywords

heart valve 5-hydroxytryptamine (5HT) receptor immunohistochemistry TaqMan cynomolgus monkey Sprague-Dawley rat

Valvular heart disease (VHD) associated with the activation of 5-hydroxytryptamine (5HT) receptors and/or increased circulating 5HT levels has been described in humans with carcinoid tumors and use of 5HT2B agonists, such as fenfluramine (Pondimin), dexfenfluramine (Redux), and ergot alkaloids (ergotamine, dihydroergotamine, and methysergide) (Connolly et al. 1997; Jick et al. 1998; Fitzgerald et al. 2000; Rothman et al. 2000; Rajamannan et al. 2001; Weissman 2001; Nebigil and Maroteaux 2003). A similar association between pergolide (Permax) and VHD has been reported, also proposed to be mediated via the 5HT2B receptor (5HT2BR) (Pritchett et al. 2002). 5HT2BR stimulation is known to cause fibroblast mitogenesis and is likely to be associated with the development of drug-induced VHD (Roy et al. 2000; Setola et al. 2003). As a consequence, 5HT2BR expression in heart valves or their interstitial cells has been the subject of several recent investigations. Fenfluramine-associated VHD is characterized by segmental or nodular thickening of affected valves, and thickening was primarily due to subendocardial fibromyxoid tissue consisting of interstitial cells and increased levels of extracellular matrix (Steffee et al. 1999). Similar microscopic finding has been described in resected valves harvested from patients undergoing long-term administration of certain ergot derivatives (e.g., ergotamine and methysergide) and from those suffering from carcinoid syndrome (Steffee et al. 1999).

The biological actions of 5HT are mediated via numerous cognate receptors. A total of 15 receptor subtypes have been discovered to date, and these subtypes belong to the following four classes of receptors: 5HT1/5, 5HT2, 5HT3, and 5HT4/6/7 (Hoyer et al. 2002). All serotonin receptors, with the exception of 5HT3, belong to the superfamily of G-protein-coupled receptors. Previous studies have reported that 5HT1B/1CR and 5HT2A/2BR subtypes are expressed in human and porcine aortic and mitral valve cells as well as in canine aortic valves (Fitzgerald et al. 2000; Roy et al. 2000). 5HT2BR is expressed in the adult cardiovascular system, as well as in the gut and brain, and in vitro mediates 5HT-induced mitogenesis in transfected fibroblasts by recruiting c-Src for cell-cycle progression via the mitogen-activated protein kinase (MAPK ERK1/2) pathway (Nebigil et al. 2000; Nebigil and Maroteaux 2003). Norfenfluramine (main metabolite of fenfluramine)-mediated VHD was recently reported to be mediated predominantly by 5HT2AR via ERK1/2 activation through both a protein kinase C (PKC) and Src stimulation (Xu et al. 2002). It has also been shown that 5HT may contribute to the development and progression of cardiac valve disease through upregulation of TGF-β via a G-protein signal transduction (Jian et al. 2002).

To elucidate the mechanistic link between 5HT receptors and VHD, it would help to define the expression of 5HT receptors in various heart valves, particularly the expression of 5HT2R in normal heart valves. Although the presence of 5HT receptors has been reported in the cardiovascular system, including vascular endothelial cells and cardiomyocytes in rat (Baxter et al. 1995; Ullmer et al. 1995) and mouse (Choi and Maroteaux 1996), detailed expression patterns for 5HT1BR, 5HT2BR, and 5-HT2CR have not been previously described in four heart valves, particularly in Sprague-Dawley (S-D) rats. To the best of our knowledge, the 5HT-receptor expression pattern has not been previously described in the heart valves of cynomolgus (CM) monkeys. In rats, where individual valve leaflets are small, the precise collection of each leaflet is often difficult and requires microsurgical resection. The fast and precise dissection possible with laser capture microdissection (LCM), combined with the ability to readily confirm the nature of the captured material, are obvious advantages of this approach, and LCM is emerging as one of the methods of choice. In this study we used real-time polymerase chain reaction (RT-PCR) (TaqMan) to quantify transcript levels of 5HT2B, 5HT2C, and 5HT1B receptors and immunohistochemistry (IHC) to detect tissue localization of these receptors in normal heart valves of male CM monkeys and S-D rats.

Materials and Methods

Four heart valves, namely, mitral, tricuspid, aortic, and pulmonary valves, were collected at necropsy from two CM monkeys (5 and 8 years of age) and eight S-D male rats (21 weeks of age) (Charles River Laboratories; Raleigh, NC). Valve leaflets were quickly snap-frozen in liquid nitrogen for TaqMan and fixed in 10% neutral buffered formalin (NBF) for IHC. For rat, valve leaflets were microsurgically resected. Because of the small sample size, individual leaflets of each heart valve were pooled from four rats for RT-PCR (TaqMan), and hearts from another six rats were fixed intact in 10% NBF and infused with OCT (optimal cutting temperature; Tissue-Tek, Torrance, CA) media for IHC and LCM/TaqMan, respectively.

RT-PCR (TaqMan) Surgically Resected Heart Valves

CM Monkey Valves. Total RNA was extracted using Tri-Reagent (Molecular Research Center; Cincinnati, OH) according to the manufacturer's recommended protocol. All RNA samples were initially quantitated via spectrophotometer prior to DNasing. Sample integrity was then checked using a 1.25% agarose gel in which 0.5 μg of RNA per sample was run in 1X [N-morpholino] propane sulfonic acid (MOPS) solution and stained with ethidium bromide. The samples were then DNased using DNA-free DNase 1 (Ambion; Austin, TX) according to the manufacturer's recommended protocol. Residual genomic DNA contamination was assessed through a minus reverse transcriptase PCR for each RNA sample. If the sample had a cycle threshold (Ct) of 38 or greater, it was deemed free of genomic contamination. If the sample had a Ct less than 38, the DNase procedure was repeated. Once the samples were cleaned of genomic DNA, they were quantitated by RiboGreen (Molecular Probes; Eugene, OR). Two μg of each RNA was converted to cDNA using random hexamers and MuLV reverse transcriptase (Applied Biosystems; Foster City, CA) according to the manufacturer's recommended protocol. cDNA (12.5 ng) from each sample was then loaded into 384-well PCR-run plates using a BioMek FX (Beckman Coulter Inc.; Fullerton, CA).

S-D Rat Heart Valves. Total RNA was extracted from four pooled samples of heart valves (mitral, tricuspid, aortic, pulmonary) using RNeasy Mini Kit (Qiagen; Hilden, Germany) according to the manufacturer's protocol. RNA quality was checked using an Agilent 2100 Bioanalyzer (Agilent Technologies; Palo Alto, CA). RNA samples were DNase treated using DNase-Free Kit (Ambion) according to the manufacturer's protocol. RNA sample was converted to cDNA using the ABI High Capacity cDNA Archive Kit (Applied Biosystems). cDNA was quantitated by spectrophotometer, diluted to 30 ng/μl, and run with rat 5HT1BR, 5HT2BR, and 5HT2CR primers ordered from the ABI TaqMan Gene Expression Assay (Applied Biosystems) service.

Immunohistochemistry

Following formalin fixation, valve samples were processed into paraffin blocks and 4-μM sections were prepared for IHC. IHC was performed using a Discovery automated staining system (Ventana Medical; Tucson, AZ). An automated IHC protocol using antibodies to the 5HT receptors (catalog #IMG-366 used at 1 μg/ml 5HT1BR; BioCarta US, Carlsbad, CA), (catalog #556334 used at 4 μg/ml 5HT2BR; BD Pharmingen, San Diego, CA), (catalog #556335 used at 4 μg/ml 5HT2CR; BD Pharmingen) was optimized using rat brain or stomach as positive control tissue (data not shown) before analyzing the valve samples. Isotype negative controls were included with each run (rabbit IgG for 5HT1BR, mouse IgG for 5HT2BR and 5HT2CR). The following steps were programmed into the automated protocol: slide deparaffinization, high pH epitope recovery, protein block, endogenous peroxidase block, primary and secondary antibody (biotinylated species-specific secondary antibodies—antimouse IgG secondary antibody was adsorbed against rat IgG) incubation, endogenous biotin block, detection (streptavidin conjugated to horseradish peroxidase), and antigen visualization using DAB as the chromagen. Following the automated protocol, sections were counterstained off-line in hematoxylin and coverslipped using a synthetic mounting medium.

Laser Capture Microdissection and TaqMan

Whole rat hearts were immediately excised, and chambers were infused with OCT media using a 10-ml syringe. Hearts were then placed in cryomold containing OCT and frozen at −80C. Seven-μm frozen tissue sections were cut onto the Histogene LCM glass slides (Arcturus Engineering; Mountain View, CA) using a Leica CM3050 cryotome (Leica Microsystems; Bannockburn, IL), and ∼15–20 slides were cut for each of the aortic, mitral, pulmonary, and tricuspid valves. Slides were stained using HistoGene LCM Frozen Section Staining Kit (Arcturus) according to manufacturer's protocol. Each set of valves was identified and microdissected using the Autopix LCM system (Arcturus) onto Macro LCM caps. The procedure took no more than 10 min per section, and the microdissected samples were immediately taken for total RNA extraction. RNA was extracted from each cap using the PicoPure RNA Isolation Kit (Arcturus) according to protocol. Recovered RNA from each set of valves was then pooled together and subjected to two rounds of amplification using RiboAmp (Arcturus) protocol. One μl of amplified RNA was then run on the Agilent Bioanalyzer (Agilent Technologies) to check for amplified product. Amplified RNA was then converted to cDNA using the ABI High Capacity cDNA Archiving Kit (Applied Biosystems) according to the manufacturer's protocol. Converted cDNA was then diluted 1:10 and run on ABI's SDS 7900HT using rat 5HTR1BR, 5HTR2BR, and 5HTR2CR primers ordered from the ABI TaqMan Gene Expression Assay service (Applied Biosystems). Results were normalized to 18S.

Figure 1.

Mitral valve from a cynomolgus (CM) monkey. Note 5HT2BR-positive interstitial cells (arrows) (X400).

Figure 2.

Aortic valve from a Sprague-Dawley (S-D) rat. Note 5HT2BR-positive interstitial cells (arrows) (X400).

Results

IHC performed on heart-valve sections revealed a strong specific expression of 5HT receptors in valvular interstitial and endothelial cells. In CM monkeys and S-D rats, mitral, tricuspid, aortic, and pulmonary valves were positive for 5HT1BR and 5HT2BR (Figures 1 and 2). The cell types showing positive staining were interstitial cells and endothelial cells lining the valvular leaflet for both species. 5HT2CR was expressed only in the endothelial cells of CM monkeys, whereas this receptor was negative in S-D rats. In both species, IHC results were correlated with 5HT2BR and 5HT1BR transcripts (mean copies per 50 ng of total RNA) in all four valves. However, mean 5HT2CR transcripts for all four valves were lower than 5HT2BR or 5HT1BR in CM monkeys (Figure 3). We also compared transcript levels of 5HT receptors among mitral, tricuspid, aortic, and pulmonary valve leaflets and saw differences in 5HT receptor transcript levels in CM monkeys. Compared with mitral, pulmonary, and tricuspid valves, the aortic valve had higher transcript levels of 5HT1BR, 5HT2BR, and 5HT2CR in CM monkeys (Figure 3). Furthermore, mean 5HT1BR transcripts were relatively higher than the 5HT2BR or 5HT2CR in the aortic valve of CM monkeys.

Because of the small size of valves in S-D rats, two sampling techniques were used: microsurgical resection and LCM. Both techniques showed the expression of 5HT2BR and 5HT1BR transcripts (mean copies per 50 ng of total RNA) in mitral, tricuspid, aortic, and pulmonary valves, whereas 5HT2CR transcripts were below the level of detection. 5HT-receptor expression patterns were comparable between techniques. In both techniques, pulmonary valves had higher 5HT2BR transcripts than aortic, tricuspid, and mitral valves. In contrast, pulmonary valves had lower 5HT1BR transcripts than those of mitral, tricuspid, and aortic valves (Figure 4A).

Discussion

The present study examined the expression of 5HT2BR, 5HT1BR, and 5HT2CR in the heart valves of CM monkeys and S-D rats. Our IHC data showed the expression of 5HT2BR, 5HT1BR, and 5HT2CR in valvular cells, but the immunostaining patterns differed among the three 5HT receptors. Cell types showing positive immunostaining for 5HT2BR and 5HT1BR were endothelial and valvular interstitial cells, whereas 5HT2CR was positive only in endothelial cells lining the valvular leaflet in both species. Although the presence of 5HT receptors has been reported in the cardiovascular system, including vascular endothelial cells and cardiomyocytes in rat (Baxter et al. 1995; Ullmer et al. 1995) and mouse (Choi and Maroteaux 1996), detailed expression patterns for 5HT1BR, 5HT2BR, and 5HT2CR have not been previously described in four heart valves of S-D rats. To the best of our knowledge, the 5HT-receptor expression pattern has not been previously described in the individual valves of CM monkeys. Previous studies have reported that 5HT1B/1C and 5HT2A/2B receptor subtypes are expressed in human and porcine aortic and mitral valve cells, as well as in canine aortic valves (Fitzgerald et al. 2000; Roy et al. 2000).

Figure 3.

5HT1B, 5HT2B, and 5HT2C receptor transcripts in the heart valves of CM monkeys. Data are shown as mean copies per 50 ng of total RNA (y-axis) for mitral, aortic, tricuspid, and pulmonary valves.

In both species, IHC results were correlated with 5HT2B and 5HT1B receptor transcripts in all four valves. However, 5HT2CR transcripts were lower than 5HT2BR or 5HT1BR in all CM monkey valves, and 5HT2CR transcripts were below the level of detection in any of the S-D rat valves. The present findings in CM monkey valves are in agreement with those reported in porcine and human aortic valves that 5HT2BR and 5HT2AR transcripts were 757- and 375-fold higher than 5HT2CR levels, respectively (Fitzgerald et al. 2000). The low 5HT2CR expression in CM monkey valves correlated with the IHC findings that positive immunostaining was confined to endothelial cells lining the valvular leaflet. The most prevalent cell type in the heart valve is interstitial cell, and interstitial cells are serotonin responsive. Thus it is likely that endothelial-specific expression and absence of expression in interstitial cells may explain, at least in part, the low transcript levels of 5HT2CR in this study. Furthermore, 5HT receptor transcript levels were variable among four heart valves in both species. Compared with mitral, pulmonary, and tricuspid valves, the aortic valve had higher transcript levels of 5HT1BR, 5HT2BR, and 5HT2CR in CM monkeys. In S-D rats, pulmonary valves had higher 5HT2BR and lower 5HT1BR transcripts than aortic, tricuspid, and mitral valves. However, it is not known what cell types reflect the 5HT-receptor transcript expression pattern in the rat heart valves because RNA from whole-valve leaflet was used. Furthermore, identification and separation of valvular interstitial cell types were not performed in this study. Taylor et al. (2000) concluded that the identification and separation of valvular interstitial cell types solely by IHC was difficult because antibodies that can specifically distinguish fibroblasts, myofibroblasts, and smooth muscle cells have not been identified.

Figure 4.

5HT1B and 5HT2B receptor transcripts in the heart valves of S-D rats. (A) TaqMan data from two sampling techniques (surgical resection and LCM) are shown as mean copies per 50 ng of total RNA (y-axis) for mitral, aortic, tricuspid, and pulmonary valves. (B,C) Amplification plots from surgical resection and LCM.

Because of the small size of rats, harvesting of individual valve leaflets is difficult, often time consuming, and requires microsurgical resection procedure and technical expertise. On the contrary, fast and precise dissection of each individual valve leaflet is possible with LCM, and the ability to readily confirm the nature of the captured material is an obvious advantage of this method. LCM is now well established as a tool for facilitating the enrichment of cells of interest from tissue sections, overcoming the problem of tissue heterogeneity. Using LCM, we were able to separately microdissect out individual valve leaflets and, with TaqMan, measure 5HT-receptor transcripts from each valve leaflet. We observed the expression of 5HT2BR and 5HT1BR transcripts in mitral, tricuspid, aortic, and pulmonary valves, whereas 5HT2C transcripts were below the level of detection. Our transcript data showed differences in terms of transcript size between these techniques (LCM and microsurgical resection); however, expression patterns were comparable. Inter-valvular differences in transcript size were also noted in both techniques; pulmonary valves had higher 5HT2BR transcripts than aortic, tricuspid, and mitral valves. These differences in expression of 5HT receptors may play a role in susceptibility or predilection toward a specific heart valve. Fenfluramine-associated VHD has been shown to have a marked predilection toward involvement of mitral and aortic valves (Graham and Green 1997; Kurz and Van Ermen 1997; Rasmussen et al. 1997; Steffee et al. 1999).

5HT2BR is expressed in the adult cardiovascular system, as well as in the gut and brain, and mediates 5HT-induced mitogenesis in transfected fibroblasts by recruiting c-Src for cell-cycle progression via the mitogen-activated protein kinase (MAPK ERK1/2) pathway (Nebigil et al. 2000; Nebigil and Maroteaux 2003). Recent findings have implicated the 5HT2BR in mediating VHD in humans with carcinoid tumors and use of 5HT2B agonists such as fenfluramine, dexfenfluramine, ergotamine, dihydroergotamine, methysergide (Connolly et al. 1997; Jick et al. 1998; Fitzgerald et al. 2000; Rothman et al. 2000; Rajamannan et al. 2001; Weissman 2001; Nebigil and Maroteaux 2003), and pergolide. MDMA (3,4-methylenedioxymethamphetamine) and its metabolite MDA (3,4-methylene-dioxyamphetamine), like fenfluramine and its N-deethylated metabolite norfenfluramine, have been shown to elicit prolonged mitogenic responses in human valvular interstitial cells via activation of 5HT2B receptors (Setola et al. 2003). These findings demonstrate the necessity of screening for 5HT2BR agonists before their therapeutic use in humans (Rothman et al. 2000; Setola et al. 2003). We demonstrated the immunostaining and quantitative transcript levels of 5HT1BR, 5HT2BR, and 5HT2CR in various heart valves of CM monkeys and S-D rats. The 5HT-receptor expression is comparable among monkeys, rats, and humans; therefore, the potential exists to gain mechanistic insight to VHD from studies in animals.

In conclusion, we have shown that the combination of IHC, LCM, and TaqMan is a useful research tool for studying 5HT-receptor expression in various heart valves. In S-D rats and CM monkeys, positive immunostaining was noted for 5HT1BR, 5HT2BR, and 5HT2CR in mitral, tricuspid, aortic, and pulmonary valves. 5HT2BR and 5HT1BR were expressed in both endothelial cells as well as valvular interstitial cells, whereas 5HT2CR was expressed only in the endothelial cells lining the valvular leaflet. In both species, IHC results were correlated with 5HT2BR and 5HT1BR transcripts (mean copies per 50 ng of total RNA) for all four valves. However, 5HT2CR transcripts were lower than 5HT2BR or 5HT1BR in all CM monkey valves, whereas 5HT2CR transcripts were below the level of detection in any of the S-D rat valves. Our data showed the expression of 5HT2BR, 5HT1BR, and 5HT2CR in the normal heart valves of CM monkeys and S-D rats. 5HT receptor IHC and TaqMan techniques may be used to study the potential mechanism of compounds with 5HT2BR agonist activity.

References

Baxter

Kennett

Blaney

Blackburn

(1995) 5-HT2 receptor subtypes: a family re-united? Trends Pharmacol Sci 16: 105–110

Choi

D-S

Maroteaux

(1996) Immunohistochemical localisation of the serotonin 5-HT2B receptor in mouse gut, cardiovascular system, and brain. FEBS Lett 391: 45–51

Connolly

Crary

McGoon

Hensrud

Edwards

Schaff

(1997) Valvular heart disease associated with fenfluraminephentermine. N Engl J Med 337: 581–588

Fitzgerald

Burn

Brown

Patterson

Corjay

Valentine

Sun

(2000) Possible role of valvular serotonin 5-HT(2B) receptors in the cardiopathy associated with fenfluramine. Mol Pharmacol 57: 75–81

Graham

Green

(1997) Further cases of valvular heart disease associated with fenfluramine-phentermine. N Engl J Med 337: 635

Hoyer

Hannon

Martin

(2002) Molecular, pharmacological and functional diversity of 5-HT receptors. Pharmacol Biochem Behav 71: 533–554

Jian

Savani

Narula

Liang

Levy

(2002) Serotonin mechanisms in heart valve disease. I. Serotonin-induced upregulation of TGF-β1 via G-protein signal transduction in aortic valve interstitial cells. Am J Pathol 161: 2111–2121

Jick

Vasilakis

Weinrauch

Meier

Jick

Derby

(1998) A population-based study of appetite-suppressant drugs and the risk of cardiac-valve regurgitation. N Engl J Med 339: 719–724

Kurz

Van Ermen

(1997) Valvular heart disease associated with fenfluramine-phentermine. N Engl J Med 337: 1772–1773

10.

Nebigil

Launay

Hickel

Tournois

Maroteaux

(2000) 5-Hydroxytryptamine 2B receptor regulates cell-cycle progression: cross talk with tyrosine kinase pathways. Proc Natl Acad Sci USA 97: 2591–2596

11.

Nebigil

Maroteaux

(2003) Functional consequence of serotonin/5-HT2B receptor signaling in heart. Role of mitochondria in transition between hypertrophy and heart failure? Circulation 108: 902–908

12.

Pritchett

Morrison

Edwards

Schaff

Connolly

Espinosa

(2002) Valvular heart disease in patients taking pergolide. Mayo Clin Proc 77: 1280–1286

13.

Rajamannan

Caplice

Anthikad

Sebo

Orszulak

Edwards

Tajik

(2001) Cell proliferation in carcinoid valve disease: a mechanism for serotonin effects. J Heart Valve Dis 10: 827–831

14.

Rasmussen

Corya

Glassman

(1997) Valvular heart disease associated with fenfluramine-phentermine. N Engl J Med 337: 1773

15.

Rothman

Michael

Baumann

Savage

Rauser

McBride

Hufeisen

(2000) Evidence for possible involvement of 5-HT_2B receptors in the cardiac valvulopathy associated with fenfluramine and other serotonergic medications. Circulation 102: 2836–2841

16.

Roy

Brand

Yacoub

(2000) Expression of 5-hydroxytryptamine receptor subtype messenger RNA in interstitial cells from human heart valves. J Heart Valve Dis 9: 256–260

17.

Setola

Hufeisen

Grande-Allen

Vesely

Glennon

Blough

Rothman

(2003) 3,4-Methylenedioxymethamphetamine (MDMA, “Ecstasy”) induces fenfluramine-like proliferative actions on human cardiac valvular interstitial cells in vitro. Mol Pharmacol 63: 1223–1229

18.

Steffee

Singh

Chitwood

(1999) Histologic changes in three explanted native cardiac valves following use of fenfluramines. Cardiovasc Pathol 8: 245–253

19.

Taylor

Allen

Yacoub

(2000) Phenotypic and functional characterization of interstitial cells from human heart valves, pericardium and skin. J Heart Valve Dis 9: 150–158

20.

Ullmer

Schmuck

Kalkman

Lübbert

(1995) Expression of serotonin receptor mRNA in blood vessels. FEBS Lett 370: 215–221

21.

Weissman

(2001) Appetite suppressants and valvular heart disease. Am J Med Sci 321: 285–291

22.

Jian

Chu

Dunlop

Rosenzweig-Lipson

(2002) Serotonin mechanisms in heart valve disease. II. The 5-HT2A receptor and its signaling pathway in cultured aortic valve interstitial cells. Am J Pathol 161: 2209–2218