© 1998 by European Society of Cardiology
Copyright © 1998, European Society of Cardiology
Potential paracrine role of the pericardium in the regulation of cardiac function
aThe Pulmonary Anesthesiology Laboratory, Department of Anesthesiology and Critical Care Medicine, Johns Hopkins Medical Institutions, Baltimore, MD 21287-3711, USA
bLaboratory of Cardiovascular Science, Gerontology Research Center, National Institute on Aging, National Institutes of Health, Baltimore, MD, USA
cDepartment of Cardiology, University of Wales College of Medicine, Cardiff, UK
dLaboratoire de Biochimie, Faculté de Médecine La Pitié-Salpétrière, Paris, France
eDepartment of Anesthesiology and Critical Care Medicine, Hôpital Lariboisière, AP-HP, Paris, France
* Corresponding author. Tel.: +33 (1) 4995 8085; Fax: +33 (1) 4995 8543; E-mail: amebazaa.lariboisiere@invivo.edu
Received 29 October 1997; accepted 3 April 1998
 |
Abstract |
|---|
 Top Abstract 1 Introduction2 Methods3 Results4 DiscussionReferences |
|---|
Objective: Both coronary and endocardial endothelium regulate cardiac contractile function via paracrine pathways. We investigated whether pericardial fluid (PF) and pericardial mesothelial cells (PMC) could exert a similar paracrine action. Methods: Both PF and PMC were extracted from sheep pericardial space. Endothelin-1, prostaglandins and atrial natriuretic factor were measured in PF in vivo. In the other hand, PMC were grown on T-75 flasks and microcarrier beads to investigate endothelin-1, nitric oxide and prostaglandin pathways in vitro. In addition, effects of PF and PMC effluent were tested on adult rat cardiac myocyte contraction in vitro. Results: In vitro, cultured PMC expressed endothelin-1 mRNA but not the endothelial nitric oxide synthase III, and released endothelin-1 and prostaglandins. Both PF and cultured PMC superfusate induced a potent, rapidly reversible decrease in the shortening of isolated rat cardiac myocytes. This effect was not associated with changes in intracellular calcium. In vivo, prostaglandins, atrial natriuretic factor and endothelin were present in PF. A greater concentration of atrial natriuretic factor was present in PF than in serum, suggesting molecular diffusion from the myocardium to PF. Preliminary results show that the instillation of vasoactive agents into the pericardial space of dogs rapidly alter coronary and systemic vascular tone, consistent with a molecular diffusion of these substances from PF into the myocardium and circulation. Conclusions: In addition to its mechanical role, the pericardium may contribute to the integration and the regulation of cardiovascular function via a paracrine mechanism.
KEYWORDS Atrial natriuretic factor; Endothelin; Mesothelium; Nitric oxide; Prostaglandins
|
1 Introduction |
|---|
|
TopAbstract1 Introduction2 Methods3 Results4 DiscussionReferences |
|---|
The pericardium and pericardial space are considered to serve a merely mechanical role in regulating cardiovascular function. Pericardial fluid (PF), present in the pericardial space, serves as a lubricant while the pericardium itself has important properties in constraining the myocardium [1]. PF is an ultrafiltrate of plasma enriched in molecules, such as prostaglandins (PG), atrial natriuretic factor (ANF) and growth factors, produced by normal and diseased hearts [2–4]. In humans, we observed that the potent cardiac growth factor, basic fibroblast growth factor (FGF-2), is more concentrated in PF than in plasma [3]. In addition, the concentration of FGF-2 is greater in PF of patients with ischemic heart disease than in patients with nonischemic heart disease [2].
Several studies have established that endocardium and coronary vascular endothelium regulate myocardial contractile function by the release of cardioactive mediators (reviewed in Refs. [5, 6]). These mediators include nitric oxide (NO), endothelin-1 (ET-1), prostanoids and a number of other as yet nonidentified substances. We hypothesized that pericardium, and particularly pericardial mesothelial cells (PMC), may affect myocardial function by a similar paracrine pathway. Cultured PMC have been described to release prostanoids [7], but more importantly to alter the phenotype and function of ventricular myocytes [8]. In coculture, PMC increased 
-myosin heavy chain isoform transcription and altered the contractile properties of isolated cardiac myocytes. The agents involved in this cell–cell cross-talk remain unidentified.
Accordingly, we investigated, in female sheep, whether (1) the PF contains and (2) cultured PMC release mediators of cardiovascular function and (3) whether alteration of the contents of the pericardial space alters cardiovascular function
|
2 Methods |
|---|
|
TopAbstract1 Introduction2 Methods3 Results4 DiscussionReferences |
|---|
2.1 PF
2.1.1 Bioactive mediators in PF
All animals were studied in accordance with institutional Animal Care and Use Committee regulations. Six month old female sheep were anesthetized with ketamine (50 mg/kg i.m.). Additional doses were given intravenously as needed to maintain anesthesia. Sheep were intubated and artificially ventilated with air containing additional oxygen to maintain PaO2150 mmHg. They were placed on a heating pad to maintain rectal temperature at 38°C. Fluid-filled catheters were inserted in the right femoral artery and in the right jugular vein to monitor arterial and right atrial pressures. A femoral venous catheter was used for volume loading. Heart rate was measured from a three-lead electrocardiograph. The chest was opened through a median sternotomy and a 0.5 cm incision was made in the anterior surface of the parietal pericardium. The pericardial space was gently washed twice with 15 ml of Hank's balanced salt solution (HBSS, composition in mM: NaCl 138.0, KCl 5.0, NaHCO3 4.0, CaCl2 1.3, MgCl2·6H2O 0.5, MgSO4·7H2O 0.4, KH2PO4 0.3, Na2HPO4 0.3) at 37°C, before starting the experiments.
In the first set of experiments, we measured PG content in PF after three sequential interventions: (1) control; (2) increased preload by volume loading (30 ml/kg saline over 10 min) and afterload by thoracic aortic constriction with linen tape; and (3) after infusion of indomethacin (300 µM) into the pericardial space to inhibit cyclooxygenase activity. At each step and after stabilization of the hemodynamic parameters, 10 ml of HBSS (37°C) was gently instilled into the pericardial space. After 10 min PF was aspirated from the pericardial space and PG content was measured. For the last intervention the 10 ml of HBSS contained indomethacin.
In a second set of experiments, we measured ANF and ET-1 concentrations in both PF and plasma. Samples were taken simultaneously, before and after infusing saline (30 ml/kg) and during atrial pacing at 160/min. During these conditions, 10 ml of HBSS was instilled into the pericardial space. Ten min after each instillation, PF was aspirated from the pericardial space and 5 ml of blood was simultaneously withdrawn from the right atrium. Specimens were placed in chilled tubes containing 0.3 M EDTA and stored (–70°C) until ANF and ET-1 assays were performed as described in Section 2.3.
2.1.2 Effects of PF on cardiac myocyte contraction
Aliquots of PF were collected from sheep as follows. After opening the chest, the pericardial space was filled with 20 ml of 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid (HEPES)-buffered saline solution (composition in mM: 137 NaCl, 4.9 KCl, 1.2 MgSO4, 1.2 NaH2PO4, 15 glucose, 1 CaCl2 and 20 HEPES; pH 7.4; 37°C). Five min later, PF was withdrawn, corrected for temperature and pH, if necessary, and superfused over cardiac myocytes. Simultaneous changes in cardiac myocyte contraction and intracellular calcium were monitored as previously described [9, 10]. Briefly, adult Wistar rat ventricular myocytes were isolated by low-Ca2+ collagenase retrograde coronary perfusion and selected for study according to previously established criteria [11]. Myocyte suspensions were placed in a chamber on the stage of an inverted fluorescence microscope (Zeiss IM35), superfused with HEPES-buffered saline at 25°C and electrically stimulated at 0.5 Hz. Isotonic twitch amplitude of a single myocyte was monitored from a bright-field image of the cell projected onto a photodiode array with a 5 ms scan time. The amplitude of cardiac myocyte shortening=(resting length–length after electrical stimulation)/resting length is expressed as% change of resting length. For measurement of intracellular calcium, a subgroup of myocytes was loaded with indo-1-acetoxymethyl (AM) ester which was excited at 350±5 nm, every 5 ms. The 410/490 nm fluorescence emission ratio was used as an index of intracellular calcium [12]. No attempt was made to calibrate cytosolic calcium concentration because of the variable subcellular compartmentation of indo-1-AM [12]. Each PF aliquot was assayed on individual stable myocyte preparations after a control myocyte shortening and calcium transient were obtained for comparison.
2.1.3 Systemic effects of vasoactive agents infused into the pericardial space
To determine whether the addition of vasoactive agents to the pericardial space alters cardiovascular function, we performed a preliminary study on two acutely instrumented dogs in which coronary artery blood flow was more easily measured than in sheep. The pericardial incision was enlarged (4 cm) to access and dissect the left descending coronary artery. When the flow probe was in place, a catheter was inserted in the pericardial space and the margins of the incision were closed. Another flow probe was placed around the left femoral artery. Arterial and right atrial pressures and heart rate were monitored as described above. The protocol consisted of measuring hemodynamic parameters before and after injection into the pericardial space of HBSS (5 ml) alone or HBSS containing nitroglycerin (0.25 mg/ml) or epinephrine (0.05 mg/ml). These different steps were separated by 20–30 min to allow the hemodynamic variables to return to control values.
2.2 PMC
2.2.1 Cell culture
PMC were harvested from sheep pericardial space without enzymes or abrading the pericardial surface. After anesthesia and ventilation, the pericardial space was opened and filled with 20 ml of sterile Dulbecco's phosphate-buffered saline (Sigma, St. Louis, MO, USA) with added penicillin G (150 U/ml) and streptomycin (150 µg/ml). This fluid was recovered and reinjected in the pericardial space several times for 2–3 min. In this fashion, 300 ml of combined cell suspensions were harvested. These were centrifuged at 200 g for 10 min. The pellet was resuspended in complete medium [Medium 199 (Sigma) supplemented with 10% fetal bovine serum (Sigma), 10% Nu serum IV (Collaborative Research, Bedford, MA, USA), L-glutamine (2 mM) (Gibco, Grand Island, NY, USA), penicillin G (50 U/ml), streptomycin (50 µg/ml) and fungizone (250 µg/l) (Gibco)], plated onto Petri culture plates (Falcon) and incubated in 5% CO2, 95% air at 37°C. The medium was changed after 1 h to remove nonattached cells and subsequently changed every 2 days. Colonies (50–100 cells) were identified using phase-contrast microscopy. The selected colonies were transferred to 24-well plates using cloning rings (Bellco) and 0.05% trypsin/EDTA treatment (Gibco). Confluent cells were subcultured using trypsin/EDTA. After three or four passages, cells were grown either in six-well plates for pharmacological experiments or in T-75 flasks. When PMC were confluent in T-75 flasks, microcarrier beads were added to the medium and flasks were continuously agitated on a rocker platform (1.5 cycles/min). When confluent on microcarrier beads, PMC were used for perfusion experiments.
For comparison to PMC, sheep pulmonary artery and endocardial endothelial cells were harvested and cultured as described previously [9, 10, 13].
2.2.2 Immunocytochemistry
These procedures were performed on subconfluent populations of cells grown on eight-well chamber slides (Nunc, Naperville, IL, USA). Previously described immunocytochemical techniques [14–16]were used to characterize the colonies of polygonal-shaped PMC (Fig. 4). Monoclonal antibodies to cytokeratin 18 (cytoskeletal proteins of 44 000 daltons (Da) [14]) which are specific to mesothelial cells were used with indirect immunofluorescence staining (Sigma) [15]. Monoclonal antibodies to human cytokeratin (AE1 and AE3; Signet, Dedham, MA, USA) that recognizes type I and type II keratins [16]and monoclonal antibodies against -actin (Sigma) were used with immunoperoxidase staining. The uptake of DiI-acetylated low density lipoproteins (LDL; Molecular Probes, Eugene, OR, USA; 10 µg/ml) was evaluated by fluorescence microscopy as previously described [17]. Endocardial endothelial cells were used as both positive (LDL) and negative (cytokeratin 18) controls.
|
2.2.3 PG release by PMC
2.2.3.1 Pharmacological stimuli in static (no-flow) conditions
As we have previously described for endothelial cells [13], PG production was measured in the supernatants of PMC in six-well plates. For these experiments, culture media was removed and the cells were washed three times with HBSS. PG content in the supernatant was measured, following a 15 min incubation at 37°C with 2 ml of: (1) HBSS (control); (2) arachidonic acid (10 µM); (3) calcium ionophore A23187 [GenBank] (Sigma) (A23187 [GenBank] , 10 µM); or (4) A23187 [GenBank] (10 µM) preincubated for 5 min with indomethacin (30 µM). At the end of the experimental period, the cells stimulated with A23187 [GenBank] were suspended in trypsin/EDTA to assess cell number and viability (trypan blue exclusion). Supernatants were analyzed [radioimmunoassay and combined gas chromatography–mass spectrometry (GC–MS), see Section 2.3.1) for 6-keto-PGF1
(the stable metabolite of prostacyclin, PGI2) and PGE2.
2.2.3.2 Effect of flow and hypoxia
The effect of pulsatile shear stress on PMC PG release was studied using our previously described perfusion system [17]. Briefly, the perfusion system was designed to provide pulsatile flow in identical, independent cartridges (Gelman 25 mm air monitoring cassette; Gelman, Ann Arbor, MI, USA; with 0.8 µm filters) containing PMC on microcarrier beads. The perfusion rate was initiated and maintained by using a single multichannel, variable speed roller-head pump (Masterflex 7553-60, Cole-Parmer, Chicago, IL, USA). Krebs solution (composition in mM: NaCl 119.7, KCl 4.6, NaHCO3 25, CaCl2 1.2, MgCl2·6H2O 0.5, NaH2PO4 1.5, Na2HPO4 0.7, D-glucose 10) was pumped from a reservoir through gassing flasks (normoxic: 21% O2, 5% CO2, 74% N2; or hypoxic: 5% O2, 5% CO2, 90% N2) and then to the cell cartridges. Oxygen tension was 
150 Torr and 40 Torr, respectively, for the normoxic and the hypoxic solution. The temperature was maintained at 37°C in servo-controlled water baths (220 A, Fisher, Orangeburg, NY, USA). Perfusates were collected, just distal to the cell cartridges, every 20 min in tubes chilled to 2°C. In both six-well plate and pulsatile flow experiments (normoxic and hypoxic conditions) PMC viabilities were consistently greater than 93%.
2.2.4 Effects of PMC effluent on cardiac myocyte contraction
The superfusing effluents of cultured PMC were collected and tested as previously described for vascular and endocardial endothelial cells [9, 10]. Cartridges were filled with PMC on microcarrier beads (2–2.5 ml microcarrier beads=approximately 5–6x106 cells) and were continuously superfused (1 ml/min) with HEPES-buffered saline solution at 37°C. After collection, effluents were corrected for temperature and pH, if necessary, infused into the rat cardiac myocyte chamber as described above, and any resultant changes in myocyte shortening or calcium transient were monitored.
2.3 Assays
2.3.1 PG
PG were measured either directly or after lipid-extraction. Supernatants were lipid-extracted over C18 columns (Sep-Pak, Waters, Milford, MA, USA) preactivated with methanol. Recovery with this technique was greater than 90% for both radioactive and nonradioactive 6-keto-PGF1 [17]. Dried extracts and nonextracted samples were stored at –70°C.
For six-well plate and flow experiments, both 6-keto-PGF1 and PGE2 were measured by radioimmunoassay using standard techniques [17]. In PF and a subgroup of supernatants of six-well plate experiments, 6-keto-PGF1 and PGE2 concentrations were determined by combined GC–MS, as previously described [18]. Samples were extracted after addition of 3.1 and 5.2 ng of tetradeuterated (2H4) analogs of 6-keto-PGF1 and PGE2 respectively, as internal standards (Merck, St. Louis, MO, USA). The differences between the amounts of prostanoid added and the amounts measured were typically less than 19%, and the intraassay coefficients of variation for repeated measurements of a single sample were less than 20%.
Prostanoid production was expressed as pg/million cells/15 min for the six-well plate experiments, pg/mg protein/min for the flow experiments and ng/10 min (production in pericardial space) for the in vivo experiments.
2.3.2 ET-1 synthesis and release, and ANF
ET-1 mRNA expression and total cellular RNA were measured in cultured PMC using the guanidinium thiocyanate/phenol/chloroform method previously described [9]. Briefly, RNA (20 µg) was separated by electrophoresis on 1% agarose gel and transferred to nylon membrane (Genescreen plus). Hybridization was carried out in hybridization buffer: [50% formamide, 5xSSC, 2.5 Denhart's solution and 1% sodium dodecyl sulfate (SDS)], 0.2 mg/ml salmon sperm single strand DNA and 1 mM EDTA, containing 32P-labeled cDNA probes complementary to sheep ET-1 mRNA (gift from J. Conary and T. Quertermous, Vanderbilt University, Nashville, TN, USA). Endocardial endothelial cell RNA was studied in the same manner, as positive control. Each blot was reprobed with a mouse tubulin riboprobe (gift from N.B. Raj, Johns Hopkins University, Baltimore, MD, USA) generated with SP6 RNA polymerase [19].
Confluent PMC were incubated for 60 min at 37°C with 2 ml of HBSS in six-well plates. The supernatants were assayed for ET-1 production. PF and plasma (collected in vivo as described above) were assayed for both ET-1 and ANF content. All samples were extracted through preactivated Sep-Pak C18 columns. Immunoreactive (ir) ANF and ET-1 were measured by radioimmunoassay using commercially available kits: ANF (Peninsula Laboratories, Belmont, CA, USA), ET-1 (NEN, Boston, MA, USA).
2.3.3 Immunoblotting of NO synthase (NOS) III
Cell proteins were separated on 10% SDS-polyacrylamide and then electrophoretically transferred to nitrocellulose membranes (HYBOND-C, Amersham, Arlington Heights, IL, USA; 90 µg protein/lane) using a Transphor electroblotter unit (Hoefer, San Francisco, CA, USA) at 100 V for 2 h. The membranes were incubated for 2 h in blocking solution containing 50 mM Tris base (pH 7.6), 4 mM MgCl2 and 14 mM NaCl (TBS) supplemented with 30 mg/ml bovine serum albumin, 0.1% Tween-20 and 2% sodium azide, and were washed twice with TBS containing 0.1% (v/v) Tween-20 and 2% (v/v) NP-40. The membranes were incubated with polyclonal antiNOS III antibodies (Transduction Laboratories, Lexington, KY, USA [20, 21]) (1:250) in TBS–Tween-20 for 1 h (similar results were obtained with monoclonal antiNOS III antibodies, data not shown). After washing the blots in TBS, 0.4% (w/v) SDS and 4% (v/v) NP-40, they were incubated with secondary antibody (antirabbit IgG, peroxidase conjugate, 1:2000 dilution) for 1 h. The blots were finally washed twice for 30 min in TBS containing 0.4% (w/v) SDS, 4% (v/v) NP-40, and processed using the enhanced chemiluminescence (ECL) detection system (Amersham), as described by the manufacturer.
2.4 Statistics
Data are presented as mean±standard error of the mean. An unpaired Student's t-test was used to compare in vitro effects of pharmacological stimuli. A two-way analysis of variance was used to analyze the effects of flow and hypoxia. Within-group comparisons were made by paired Student's t-test on absolute data. A P<0.05 was considered significant.
|
3 Results |
|---|
|
TopAbstract1 Introduction2 Methods3 Results4 DiscussionReferences |
|---|
3.1 Mediators in the PF
Table 1A shows that under basal conditions, PGI2 content was sevenfold greater than PGE2 content in PF in vivo. Acute volume loading and aortic constriction increased right atrial pressure from 3 to 14 mmHg and induced a threefold increase in PGI2 and PGE2 content in PF. To assess whether the PG present in the PF were newly synthesized, indomethacin was added to PF. This resulted in an immediate decrease in PGI2 and PGE2 content in the PF.
|
Fig. 1 shows that, under basal conditions, ANF concentration in PF was fourfold greater than in plasma. Atrial stretch and tachycardia increased ANF concentrations in both plasma and PF, but ANF concentrations in PF remained greater than in plasma. By contrast, ET-1 concentrations in PF did not respond to atrial stretch and tachycardia. PF ET-1 concentration was consistently lower than in plasma during both basal conditions and atrial dilation and tachycardia (Fig. 1).
|
, P<0.05 indicates the comparison between control and volume loading and tachycardia for the same type of fluid (paired Student's t-test).3.2 Morphological and immunological characteristics of PMC in vitro (Fig. 2)
Observation of PMC in monolayer culture by phase-contrast light microscopy showed cells with a polygonal shape at confluence and a pavement-like appearance, as previously described [7]. The positive staining observed with monoclonal antibodies to cytokeratin 18 and its perinuclear pattern are characteristic of mesothelial cells [15]. Abundant keratin AE1 and AE3 were also demonstrated in confluent and subconfluent PMC. These data are consistent with reported cytokeratin patterns in rat and rabbit mesothelial cells [15, 22]. In addition, PMC never stained for
-actin and showed no detectable uptake of DiI-acetylated-LDL. Using similar techniques, sheep endocardial endothelial cells did not contain cytokeratin 18 nor AE1/AE3 but demonstrated uptake of DiI-acetylated-LDL.
|
3.3 PMC release and expression of mediators
In six-well plates, PMC showed a substantial increase in PG production in response to both arachidonic acid (10 µM) and calcium ionophore A23187 [GenBank] (Table 1B). PGE2 production was greater than that of PGI2. Indomethacin inhibited A23187 [GenBank] -stimulated PGI2 and PGE2 production.
In contrast to the static experiments, during pulsatile flow conditions the production of PGI2 and PGE2 was similar (Fig. 3). During both normoxia and hypoxia, the rate of production of both PGI2 and PGE2 increased over time (P<0.001). Hypoxia, however, increased only PGI2 production (P<0.02) but not PGE2 production. Throughout the experiment, the PGI2/PGE2 ratio remained close to 1. These results were confirmed by GC–MS (data not shown).
|
PMC in culture expressed ET-1 mRNA (Fig. 4) and released ET-1 at the rate of 66.3±9.3 pg/106 cells (n=9). NOS III expression was undetectable in PMC in culture while it was present in cultured pulmonary artery and endocardial endothelial cells (Fig. 5).
|
3.4 Effects of PF and PMC superfusate on cardiac myocyte contraction (Fig. 6, Table 2)
In freshly isolated cardiac myocytes, PF rapidly (within 90 s) decreased the amplitude of myocyte shortening, increased diastolic length by 1.8±0.4 µm and induced an earlier time to peak shortening. These effects were not associated with a decrease in the indo-1 fluorescence ratio amplitude. A reduction in the amplitude of myocyte shortening without a significant decrease in the calcium transient implies a reduction in the relative myofilament response to calcium, as previously described [10, 12]. After replacing the PF with HEPES-buffered saline, myocyte shortening and indo-1 fluorescence demonstrated a rebound that lasted several minutes after washout.
|
|
PMC superfusate also decreased the amplitude of myocyte shortening and increased diastolic length by 0.9±0.1 µm. The maximal reduction in the amplitude of myocyte shortening was observed after 3.3±0.2 min. Twitch relaxation consistently occurred earlier with PMC superfusate than with HEPES, with an earlier time to peak shortening. These effects were not associated with significant changes in either systolic or diastolic indo-1 fluorescence ratio. PMC superfusate effects were rapidly reversed by perfusion with HEPES-buffered saline. The reduction in the relative myofilament response to calcium of PMC superfusate was similar to that observed with PF. Both of these resembled the previously described myofilament desensitizing activity of vascular and endocardial endothelial cell superfusate [5, 9, 10].
3.5 Effects of vasoactive agents administered in the pericardial space in vivo
Preliminary results show that HBSS instilled into the pericardial space did not alter hemodynamic parameters. In contrast, administration of nitroglycerin into the pericardial space increased both coronary and femoral blood flow and decreased right atrial pressure with no changes in heart rate and mean arterial pressure (Table 3). Epinephrine induced an increase in all measured parameters by over 50%. No hemodynamic responses occurred before 1 min. All effects occurred within 2–4 min and lasted 10–15 min. Coronary and femoral arterial flow effects occurred within the same time frame.
|
|
4 Discussion |
|---|
|
TopAbstract1 Introduction2 Methods3 Results4 DiscussionReferences |
|---|
These data demonstrate (1) the presence of cardiovascularly active substances in the pericardial space, the concentration of which changes in response to changes in cardiovascular status and (2) that PMC release biologically active cardiovascular mediators.
4.1 Cardiovascular mediators in the pericardial space: the pericardial "milieu"
PF has been described as a passive ultrafiltrate of plasma. In greyhounds, rabbits and humans, ionic composition of PF and plasma are comparable [23]. PF is enriched by the presence of PG, ANF, ET-1 and growth factors in basal conditions, in human and in dog [3, 4]. We confirm these reports and describe a response in PG, ANF and ET-1 concentration in the pericardial space in response to physiological stimuli. Changes in preload and afterload increased the PGI2 and PGE2 content of the pericardial space, an effect inhibited by indomethacin. Additionally, increased preload with atrial pacing acutely increased the pericardial space content of ANF, but not of ET-1. Finally, we have demonstrated, for the first time, that PF exerts an effect on isolated cardiac myocyte function similar to that previously described for vascular and endocardial endothelial cells [5, 10, 24].
We next determined whether vasoactive mediators placed in the pericardial space could alter systemic cardiovascular function. In a preliminary study to explore this issue, pericardial administration of nitroglycerin and epinephrine induced alterations in both coronary and femoral blood flow, heart rate and mean arterial pressure (Table 2). These observations, taken with previously described data [4, 25, 26], show that vasoactive mediators such as PG, NO and catecholamines in the pericardial space could alter both local and systemic cardiovascular function.
4.2 PMC characterization
The cells harvested from the pericardial space by our technique were morphologically similar to the PMC morphology reported by Satoh and Prescott [7]. These cells demonstrated cytokeratin 18 and the epithelial keratin AE1/AE3 labeling that both specifically identifies mesothelial cells [15, 16, 27]. Our pericardial cells showed no detectable uptake of LDL at 10 µg/ml. This result is in contrast with the data of Van Hinsbergh et al. who reported DiI-acetylated-LDL uptake by human PMC at 20 µg/ml [15]. Our data however are consistent with the finding of Satoh and Prescott who reported no detectable uptake of DiI-acetylated-LDL by bovine PMC at 10 µg/ml [7]. These differences may be related to the concentration of DiI-acetylated-LDL used and/or to the species studied. Nevertheless, several reports have established that, unlike cytokeratins, conventional histological and biochemical criteria such as phase-contrast microscopy, DiI-acetylated-LDL uptake, factor VIII-related antigen and angiotensin converting enzyme expression are not sufficient to distinguish among vascular, microvascular endothelial cells and mesothelial cells [15, 16, 27]. Thus, the uniform presence of cytokeratins in cells harvested from the pericardial space by our technique demonstrate that our cells were PMC.
4.3 Mediator release by PMC
PMC in culture mainly released PGE2 (Table 1B). Under various conditions, including pharmacological stimuli and pulsatile flow, the ratio of PGI2/PGE2 produced by PMC was consistently less than 1. These data are compatible with those previously reported with cultured mesothelial cells and with pieces of parietal pericardium in which PGI2/PGE2 ratios were usually 
1 [28, 29]. In contrast, PGI2 is the major prostanoid released by vascular and cardiac endothelial cells and endothelial PGI2/PGE2 ratios are usually greater than 10 [17]. In our study, flow and acute hypoxia stimulated cultured PMC PGI2 release whereas PGE2 release responded only to flow. Thus, PMC released vasoactive prostanoids and this release responded to pharmacologically and physiologically relevant stimuli.
In vitro, PMC expressed ET-1 mRNA and released ET-1, but did not express endothelial constitutive NOS (NOS III). The rate of ET release by sheep PMC was similar to that from rat PMC [30]and sheep cardiac endothelial cells [9]. Eid et al. have shown that in cocultures, PMC induce a marked increase in cardiomyocyte β-myosin heavy chain isoform transcription, a reduction in -actin expression and a marked alteration of cardiac myocyte contraction amplitude [8]. No agent responsible for these changes was identified. ET is a potent mitogenic factor and can alter myocyte contractility [31]. These results taken together suggest the possibility that PMC-derived ET mediates this potentially important modulation of cardiac myocyte function by PMC.
The PMC superfusate increased the resting length of adult isolated cardiac myocytes (Fig. 6, Table 2) and decreased both the amplitude of myocyte shortening and time to peak shortening. These alterations appeared rapidly within 90 s, were maximum at about 3 min and were also rapidly reversible. No significant associated changes in the intracellular calcium transient were noted. Shah et al. have presented evidence for an unidentified low molecular weight factor released by vascular and cardiac endothelial cells, which has a similar inhibitory effect on myocardial contraction and was termed "myocardial desensitizing agent" [10]. Recently, Pepper et al. reported a similar effect in the coronary effluent of isolated hearts [24]. The present results show, for the first time, that PMC in culture have an effect on cardiac myocyte contraction similar to that of this "myocardial desensitizing agent" and suggest the presence of this agent in PF in vivo.
We chose to assess the effects of PF/PMC on myocardial function using isolated cardiac myocytes. The use of pure single cell preparations (i.e. PMC/cardiac myocytes) allowed us to study the paracrine interaction between these cells in a carefully defined manner. This would be much more difficult in multicellular preparations, which contains several cell types, i.e., cardiac myocytes, microvascular and endocardial endothelial cells, and fibroblasts. A further advantage was the ability to correlate changes in cell contraction with simultaneously measured intracellular calcium concentration in the same cell. Again, this is more difficult to achieve in multicellular preparations where there may be a considerable heterogeneity in these parameters [32, 33], except in isosarcometric preparations [34]. In interpreting the observed results, however, it should be remembered that different myocardial preparations provide different information. It is well-established that myocyte shortening responses to a wide variety of interventions parallel those of force or shortening changes in intact tissue [33, 35]. Thus, it is likely that PF/PMC would also alter peak force. On the other hand, in externally unloaded myocytes, the effect of external load-dependent factors clearly cannot be addressed. In addition, the effects of possible changes in "passive" cellular elasticity (e.g. resulting from altered properties of titin or microtubules [36, 37]) will be different between unloaded myocytes and isometric preparations (single or multicellular). In unloaded cells, changes in passive properties will manifest as altered resistance to shortening and restoring force, whereas in isometric preparations changes will manifest as alterations in passive stiffness [33]. With respect to the present study, the most likely mechanism for the observed effects is a decrease in myofilament response to calcium. Calcium-mediated effects were excluded by measurement of calcium transients. Changes in passive cellular properties are also extremely unlikely because (a) the observed effects were very rapid in onset and rapidly reversible, quite unlike changes associated with microtubular disassembly [36], and (b) previous studies have shown that changes in cytoskeletal properties have minor impact on the function of healthy adult myocardium [33].
By releasing PG, ET-1 and other inotropic agent(s), PMC could alter the function of subjacent myocardium in a paracrine fashion. These mediators have been shown to directly alter myocardial myocyte function [5, 10, 17]. Additionally, the diffusion of PG and ET-1 from the pericardial space to the myocardium could also affect the myocardial circulation by a direct action on coronary vasculature. Although morphologically different, the pericardium, vascular endothelium and endocardial endothelium release similar cardioactive factors. Thus, the pericardium and the myocardial endothelium could act in concert to modulate myocardial contractile performance in a paracrine fashion by releasing these interacting cardioactive agents.
4.4 The origin of the contents of the pericardial space
The presence of active cardiovascular mediators in the PF raises the question of their origin. We have shown that PMC in vitro produce PGI2 and PGE2 in relatively equal proportions. ET-1 and the "myocardial desensitizing agent" are also produced by isolated PMC. Thus, PMC are a potential source of the cardioactive mediators found in the pericardial space in vivo. Additionally, physiologically relevant stimuli in vitro (perfusion-induced shear stress and hypoxia) alter PMC mediator production. In vivo, this could contribute to the demonstrated changes in pericardial space content following changes in myocardial loading and heart rate.
Another potential source of mediators in the pericardial space include the myocardium. We and others have recently demonstrated, that PF may reflect the composition of cardiac interstitium in normal and diseased hearts in humans [2, 3]. This is supported by the fact that PF contains molecules up to 40 kDa, including ANF [molecular weight (MW) 3 kDa] and FGF-2 (MW 18–25 kDa) that migrate from the myocardium through epicardium into the pericardial space and are more concentrated in PF than in serum [3]. In this study, we confirm, under basal conditions, that ANF concentration in fluid removed from the pericardial space was fivefold greater than in plasma. In addition, we report that volume loading and atrial tachycardia, both known stimulators of ANF production by atrial myocytes, increased the release of ANF into the pericardial space as well as increasing the plasma concentration (Fig. 1). These observations confirm a myocardial source of ANF in the PF and are consistent with molecular diffusion from the myocardium to pericardial space as previously described [38].
Analogously, other mediators of cardiovascular function found in the pericardial space could arise from the myocardium. The myocardium is a major source of PGI2 [39]. The release of PG by coronary endothelial cells is tenfold greater than that by cardiac fibroblast-like cells and cardiomyocytes [40]. In addition, we have recently shown that endocardial endothelial cells are a potent source of PGI2 [17]. PGI2 is the dominate PG found in the pericardial space. Miyazaki et al. found that the ratio of PGI2/PGE2 in dog's PF was around 6:1 under basal conditions and increased to 9:1 after instillation of arachidonic acid. The addition of indomethacin inhibited PG release [4]. In PF, in vivo, we consistently found a greater abundance of PGI2 than of PGE2 both in basal and stimulated conditions (PGI2/PGE2 ratio approximately 10:1). Because PMC in vitro produce relatively equal amounts of PGI2 and PGE2 and because we consistently found PGI2PGE2 in the PF, PMC are likely not the sole source of PGI2 in the pericardial space. Thus PGI2 in the pericardial space could originate either from the myocardial endothelial cells or alternatively from the pericardial microvascular endothelial cells [41]. The cyclooxygenase inhibitor, indomethacin, rapidly inhibited the local production of PG (Table 1). Because PF PGE2 and PGI2 content were both decreased, indomethacin must have not only acted within the pericardial space but also decreased production from cells other than the PMC such as cardiac or pericardial microvascular endothelial cells. In conclusion, these combined data are consistent with a myocardial source of PF PG in addition to those produced by PMC.
A potential source of mediators of cardiovascular function in the pericardial space is the circulating blood. However, Dusting and Nolan have shown that intravenous infusions of PGI2 at rates achieving supraphysiologic concentrations in the bloodstream were not accompanied by detectable increases of PGI2 in the PF [26]. The fact that ANF and ET-1 concentrations significantly differ between the PF and plasma is also not consistent with passive diffusion of these mediators into the PF from the blood. Thus, blood seems an unlikely source of the active mediators examined here.
In summary, our data support the notion that the pericardial space may act as a reservoir. Mediators of cardiovascular function that are released by PMC or diffuse from cardiac tissues in the pericardial space could be concentrated and subsequently released towards the myocardium and/or blood vessels. In addition, PMC may directly alter cardiac myocyte contractile function. The relative role of such regulatory pathways relative to conventional regulation by neurohumoral factors, loading conditions and the Frank–Starling mechanism, requires further study.
Time for primary review 26 days.
|
Acknowledgements |
|---|
The authors thank Drs. Walter Hubbard, Harold Spurgeon and Timothy G. Buchman, Meena Abraham and Tim Burman for their outstanding technical contribution. Ajay M. Shah is the recipient of a UK Medical Research Council Senior Clinical Fellowship. This work was supported by a NHLBI RO-1 (HL 39138-03) grant (James L. Robotham).
|
References |
|---|
|
TopAbstract1 Introduction2 Methods3 Results4 DiscussionReferences |
|---|
- Beloucif S., Takata M., Robotham J.L. Influence of pericardial constraint on atrio–ventricular interactions. Am J Physiol (1992) 263:H125–H134.[Web of Science][Medline]
- Fujita M., Ikemoto M., Kishishita M., et al. Elevated basic fibroblast growth factor in pericardial fluid of patients with unstable angina. Circulation (1996) 94:610–613.
[Abstract/Free Full Text] - Corda S., Mebazaa A., Gandolfini M., et al. Trophic effect of human pericardial fluid on adult cardiac myocytes. Differential role of fibroblast growth factor-2 and factors related to ventricular hypertrophy. Circ Res (1997) 81:680–688.
- Miyazaki T., Pride H.P., Zipes D.P. Prostaglandins in the pericardial fluid modulate neural regulation of cardiac electrophysiological properties. Circ Res (1990) 66:163–175.
[Abstract/Free Full Text] - Shah A. Paracrine modulation of heart cell function by endothelial cells. Cardiovasc Res (1996) 31:847–867.
[Abstract/Free Full Text] - Brutsaert D.L., Andries L.J. The endocardial endothelium. Am J Physiol (1992) 263:H985–H1002.[Web of Science][Medline]
- Satoh K., Prescott S.M. Culture of mesothelial cells from bovine pericardium and characterization of their arachidonate metabolism. Biochim Biophys Acta (1987) 930:283–296.[Medline]
- Eid H., Larson D.M., Springhorn J.P. Role of epicardial mesothelial cells in the modification of phenotype and function of adult rat ventricular myocytes in primary coculture. Circ Res (1992) 71:40–50.
[Abstract/Free Full Text] - Mebazaa A., Mayoux E., Maeda K., et al. Paracrine effects of endocardial endothelial cells on myocyte contraction mediated via endothelin. Am J Physiol (1993) 265:H1841–H1846.[Web of Science][Medline]
- Shah A.M., Mebazaa A., Wetzel R.C., et al. A novel cardiac myofilament desensitizing factor released by endocardial and vascular endothelial cells. Circulation (1994) 89:2492–2497.
[Abstract/Free Full Text] - Capogrossi M.C., Kort A.A., Spurgeon H.A., et al. Single adult rabbit and rat cardiac myocytes retain the Ca2+- and species-dependent systolic and diastolic properties of intact muscle. J Gen Physiol (1986) 88:589–613.
[Abstract/Free Full Text] - Spurgeon H.A., Stern M.D., Baartz G., et al. Simultaneous measurement of Ca2+, contraction, and potential in cardiac myocytes. Am J Physiol (1990) 258:H574–H586.[Web of Science][Medline]
- Mebazaa A., Martin L.D., Robotham J.L., et al. Right and left ventricular cultured endocardial endothelium produces prostacyclin and PGE2. J Mol Cell Cardiol (1993) 25:245–248.[CrossRef][Web of Science][Medline]
- Wu Y., Parker L.M., Binder N.E., et al. The mesothelial keratins: a new family of cytoskeletal proteins identified in cultured mesothelial cells and nonkeratinizing epithelia. Cell (1982) 31:693–703.[CrossRef][Web of Science][Medline]
- Van Hinsbergh V.W.M., Kooistra T., Scheffer M.A., et al. Characterization and fibrinolytic properties of human omental tissue mesothelial cells, comparison with endothelial cells. Blood (1990) 75:1490–1497.
[Abstract/Free Full Text] - Tseng S.C.G., Jarvinen M.J., Nelson W.G., et al. Correlation of specific keratins with different types of epithelial differentiation: monoclonal antibody studies. Cell (1982) 30:361–372.[CrossRef][Web of Science][Medline]
- Mebazaa A., Wetzel R.C., Cherian M., et al. Comparison between endocardial and great vessel endothelial cells: morphology, growth, and prostaglandin release. Am J Physiol (1995) 268:H250–H259.[Web of Science][Medline]
- Liu M.C., Bleecker E.R., Proud D., et al. Profiling of bisenoic prostaglandins and thromboxane B2 in bronchoalveolar fluid from the lower respiratory tract of human subjects by combined capillary gas chromatography–mass spectrometry. Prostaglandins (1988) 35:67–79.[CrossRef][Web of Science][Medline]
- Schoeniger L., Curtis W., Esnaola N., et al. Myocardial heat shock gene expression in pigs is dependent on superoxide anion generation at reperfusion. Shock (1994) 1:31–35.[Web of Science][Medline]
- Giraldez R.R., Panda A., Xia Y., et al. Decreased nitric-oxide synthase activity causes impaired endothelium-dependent relaxation in the postischemic heart. J Biol Chem (1997) 272:21420–21426.
[Abstract/Free Full Text] - Shaul P.W., Smart E.J., Robinson L.J., et al. Acylation targets endothelial nitric-oxide synthase to plasmalemmal caveolae. J Biol Chem (1996) 271:6518–6522.
[Abstract/Free Full Text] - Mackay A.M., Tracy R.P., Craighead J.E. Cytokeratin expression in rat mesothelial cells in vitro is controlled by extracellular matrix. J Cell Sci (1990) 95:97–107.
[Abstract/Free Full Text] - Gibson A.T., Segal M.B. A study of the composition of pericardial fluid, with special reference to the probable mechanism of fluid formation. J Physiol (London) (1978) 277:367–377.
[Abstract/Free Full Text] - Pepper C., Lang D., Lewis M., et al. Endothelial inhibition of myofilament calcium response in intact cardiac myocytes. Am J Physiol (1995) 269:H1538–H1544.[Web of Science][Medline]
- Miyazaki T., Zipes D.P. Pericardial prostaglandin biosynthesis prevents the increased incidence of reperfusion-induced ventricular fibrillation produced by efferent sympathetic stimulation in dogs. Circulation (1990) 82:1008–1019.
[Abstract/Free Full Text] - Dusting G.J., Nolan R.D. Stimulation of prostacyclin release from the epicardium of anaesthetized dogs. Br J Pharmacol (1981) 74:553–562.[Web of Science][Medline]
- Chung-Welch N., Patton W.F., Yen-Patton A., et al. Phenotypic comparison between mesothelial and microvascular endothelial cell lineages using conventional endothelial cell markers, cytoskeletal protein markers and in vitro assays of angiogenic potential. Differentiation (1989) 42:44–53.[CrossRef][Web of Science][Medline]
- Nolan R.D., Dusting G.J., Jakubowski J., et al. The pericardium as a source of prostacyclin in the dog, ox and rat. Prostaglandins (1982) 24:887–902.[CrossRef][Web of Science][Medline]
- Coene M.-C., Solheid C., Claeys M., et al. Prostaglandin production by cultured mesothelial cells. Arch Int Pharmacodyn (1981) 249:316–318.[Medline]
- Eid H., de Bold M., Chen J., et al. Epicardial mesothelial cells synthetize and release endothelin. J Cardiovasc Pharmacol (1994) 24:715–720.[Web of Science][Medline]
- Krämer B.K., Nishida M., Kelly R.A., et al. Endothelins. Myocardial actions of a new class of cytokines. Circulation (1992) 85:350–356.
[Abstract/Free Full Text] - Krueger J., Pollack G. Myocardial sarcomere dynamics during isometric contraction. J Physiol (1975) 251:627–643.
[Abstract/Free Full Text] - Brady A. Mechanical properties of isolated cardiac myocytes. Physiol Rev (1991) 71:413–428.
[Abstract/Free Full Text] - Janssen P., de Tombe P. Uncontrolled sarcomere length shortening increases the intracellular Ca2+ transient in rat cardiac trabeculae. Am J Physiol (1997) 272:H1892–H1897.[Web of Science][Medline]
- Shah A, Sollott S, Lakatta E. Physio–pharmacological evaluation of myocardial performance: an integrative approach. Cardiovasc Res 1998;in press.
- Tsutsui H., Ishihara K., Cooper G. Cytoskeletal role in the contractile dysfunction of hypertrophied myocardium. Science (1993) 260:682–687.
[Abstract/Free Full Text] - Labeit S., Kolmerer B., Linke W. Emerging roles in physiology and pathophysiology. Circ Res (1997) 80:290–294.
[Abstract/Free Full Text] - Page E., Upshaw-Earley J., Goings G. Permeability of rat atrial endocardium, epicardium, and myocardium to large molecules. Stretch-dependent effects. Circ Res (1992) 71:159–173.
[Abstract/Free Full Text] - De Deckere E.A.M., Nugteren D.H., Ten Hoor F. Prostacyclin is the major prostaglandin released from the isolated perfused rabbit and rat heart. Nature (1977) 268:160–163.[CrossRef][Medline]
- Linssen M.C.J.G., Engels W., Lemmens P.J.M.R., et al. Production of arachidonic acid metabolites in adult rat cardiac myocytes, endothelial cells, and fibroblast-like cells. Am J Physiol (1993) 264:H973–H982.[Web of Science][Medline]
- Amano J., Suzuki A., Sunamori M., et al. Atrial natriuretic peptide in the pericardial fluid of patients with heart disease. Clin Sci (Colch) (1993) 85:165–168.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. L. Brutsaert Cardiac Endothelial-Myocardial Signaling: Its Role in Cardiac Growth, Contractile Performance, and Rhythmicity Physiol Rev, January 1, 2003; 83(1): 59 - 115. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. R. Hermans, H. van Essen, H. A. J. Struijker-Boudier, R. M. Johnson, F. Theeuwes, and J. F. M. Smits Pharmacokinetic Advantage of Intrapericardially Applied Substances in the Rat J. Pharmacol. Exp. Ther., May 1, 2002; 301(2): 672 - 678. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. M. Arriero, J. C. de la Pinta, M. Escribano, A. Celdran, L. Munoz-Alameda, J. Garcia-Canete, A. M. Jimenez, S. Casado, J. Farre, and A. Lopez-Farre Aspirin Prevents Escherichia coli Lipopolysaccharide- and Staphylococcus aureus-Induced Downregulation of Endothelial Nitric Oxide Synthase Expression in Guinea Pig Pericardial Tissue Circ. Res., April 5, 2002; 90(6): 719 - 727. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K Tambara, M Fujita, N Nagaya, S Miyamoto, A Iwakura, K Doi, G Sakaguchi, K Nishimura, K Kangawa, and M Komeda Increased pericardial fluid concentrations of the mature form of adrenomedullin in patients with cardiac remodelling Heart, March 1, 2002; 87(3): 242 - 246. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Klemola, I. Tikkanen, O. Vuolteenaho, L. Toivonen, and M. Laine Plasma and pericardial fluid natriuretic peptide levels in postinfarction ventricular dysfunction Eur J Heart Fail, January 1, 2001; 3(1): 21 - 26. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. Arriero, J. C. de la Pinta, M. Escribano, A. Celdran, L. Munoz-Alameda, J. Garcia-Canete, A. M. Jimenez, S. Casado, J. Farre, and A. Lopez-Farre Aspirin Prevents Escherichia coli Lipopolysaccharide- and Staphylococcus aureus-Induced Downregulation of Endothelial Nitric Oxide Synthase Expression in Guinea Pig Pericardial Tissue Circ. Res., April 5, 2002; 90(6): 719 - 727. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||