Skip to main content
NCBI home page
British Journal of Pharmacology logoLink to British Journal of Pharmacology
. 1997 Feb;120(Suppl 1):494–503. doi: 10.1111/j.1476-5381.1997.tb06839.x

Bioassay of prostacyclin and endothelium-derived relaxing factor (EDRF) from porcine aortic endothelial cells

RJ Gryglewski 1,1, S Moncada 1,2, RMJ Palmer 1
PMCID: PMC3224334  PMID: 9142426

Abstract

  1. A cascade superfusion technique has been developed for the differential bioassay of prostacyclin and endothelium-derived relaxing factor (EDRF) released from porcine aortic endothelial cells cultured on microcarriers, packed into a column and perfused.

  2. Bradykinin (Bk; 20–100 Nm) released prostacyclin (9.6 ± 1.5 Nm per 106 cells; mean ± s.e.mean, n = 9) and prostaglandin E2 (PGE2; 2.1 ± 0.6 Nm per 10° cells) from the column measured by relaxation of strips of bovine coronary artery (BCA) and rabbit mesenteric or coeliac artery, respectively. The presence of these prostanoids in the effluent was confirmed by specific radioimmunoassays.

  3. A23187 (500–2000 Nm) also released both prostacyclin and PGE2 from the cells. This release was long-lasting and not reproducible.

  4. Bk (20–100 Nm) and A23187 (30–300 Nm) released EDRF from the column. This was detected in a cascade of four rabbit aortic strips (RbA), denuded of endothelium and contracted with U46619 or phenylephrine. The relaxation of the RbA strips caused by EDRF was progressively attenuated down the cascade (half-life < 7 s) and was not affected by indomethacin.

  5. EDRF and prostacyclin could be differentially bioassayed in a cascade of alternating RbAs and BCAs as prostacyclin did not relax RbAs and the time delay to the BCAs destroys EDRF. EDRF could be bioassayed on its own when the endothelial cells were treated with indomethacin.

  6. 5-Hydroxytryptamine 0.2, noradrenaline 1.0, platelet-activating factor (Paf-acether) 1.0, formylmethionyl-leucyl-phenylalanine 1.0, acetylcholine 0.5, bethanecol 0.5, adenosine diphosphate 0.25 and angiotensin II 0.1 μm did not release either prostanoids or EDRF from the column.

Introduction

Prostacyclin (Moncada et al., 1976; Gryglewski et al., 1976) and endothelium-derived relaxing factor (EDRF; Furchgott & Zawadzki, 1980) are both products of vascular endothelium which play a role in vasodilatation. Cultures of vascular endothelial or smooth muscle cells release prostacyclin and prostaglandin E2 (PGE2) either spontaneously (Ager et al., 1982) or in response to mechanical (Johnson, 1980) or chemical stimulation (Gimbrone & Alexander, 1975; Weksler et al., 1978; Hong, 1980). The development of a technique in which endothelial cells are cultured on microcarrier beads, packed into columns and perfused (Ryan et al., 1980; Davies, 1981) has allowed the study of the metabolism of arachidonic acid (AA; Pearson et al., 1983a, b; Gordon & Martin, 1983). More recently, Cocks et al. (1985) have used a similar approach to generate EDRF and detect it using a ring of canine coronary artery denuded of endothelium. We have now combined the technique of perfused cultured endothelial cells with the superfusion technique for the detection of biologically active substances (Vane, 1964) and developed a method for studying the simultaneous release and actions of both prostacyclin and EDRF by differential bioassay.

Methods

(a) Isolation and culture of porcine aortic endothelial cells

Fresh porcine thoracic aortae were obtained from an abattoir in phosphate buffered saline (PBS; pH 7.2) containing antibacterial-antimycotic mixture, and washed externally with 70% ethanol and rinsed in PBS. The lumen was then washed with PBS, the connective tissue and fat removed and the intercostal arteries ligated. The lumen was washed again and the lower end clamped prior to filling with PBS containing 1 mg ml−1 collagenase. After 15 min incubation at 37°C the solution was removed and the cells obtained washed twice by centrifugation at 180 g for 10 min in culture medium (Hepes buffered Dulbecco's modification of Eagle's medium containing 20% foetal calf serum and antibacterial-antimycotic mixture).

The endothelial cells from each aorta were suspended in culture medium and incubated separately overnight in 25 cm2 culture flasks to allow small patches of endothelial cells to adhere. The non-adherent cells were then removed by washing three times with medium and culture continued for a further 24 h to allow the cell clusters to spread.

The cells from 6–8 aortae were then removed from the culture flasks with 0.125% trypsin-versene at room temperature and washed twice in culture medium before being seeded onto 1 g hydrated Cytodex 3 microcarrier beads. The cells and beads were incubated at 37°C for 8 cycles, each consisting of 5 min stirring at 40 r.p.m. and 30 min at rest, before being continuously stirred at 40 r.p.m. in 500 ml culture medium on a microcarrier stirrer system (Techne). After 3 to 4 days in culture the cells reached confluence on the beads. This was confirmed by phase contrast microscopy. Before packing the column for perfusion experiments, the cells on 0.1 ml of beads were counted by incubation for 1 h in 1 ml of 0.1% crystal violet in 0.1 m citric acid and the nuclei released counted in a haemocytometer (Sanford et al., 1951). The volume of the microcarriers in the column was determined at the end of the experiment.

(b) Electron microscopy

Microcarriers coated with cells were fixed for light and electron microscopy in 0.1 m cacodylate buffered 3% glutaraldehyde (pH 7.2) for at least 1 h. The microcarriers were prepared for scanning electron microscopy (Malick & Wilson, 1975) and examined on a Philips 500 SEM. The cells exhibited the ‘cobblestone’ morphology and numerous microvilli characteristic of endothelial cells (Figure 1).

Figure 1.

Figure 1

Open in a new tab

Electronmicrograph of a microcarrier bead covered with endothelial cells. Bar = 10 μ.

(c) Preparation of the cell column

Between 1 and 3 ml of microcarrier beads containing 1–9 × 107 endothelial cells (5–14 day old cultures) were packed into a K16 chromatographic column, cut down to 10 cm length and maintained at 37°C (Pharmacia). The ends of the column were closed with modified flow adaptors (Pharmacia) which prevented microcarrier beads from escaping from the column (Figure 2). After packing, the column was continuously perfused with Krebs buffer gassed with a mixture of 95% O2 and 5% CO2 containing (Mm): Na+ 144.0, K+ 5.9; Ca2+ 2.5, Mg2+ 1.2, Cl− 128.7, HCO3− 25.0, SO42– 1.2, H2PO4− 1.2 and glucose 11.0. The Krebs solution was pumped through the column at 5 ml min−1 via a heated (37°C) glass coil which also served as a gas trap and as an injection port for the test substances (Figure 2). The void volume between the upper layer of the cells and the outlet of the capillary tubing leading to the superfusion cascade was 80 μl. The time taken for the superfusion fluid to reach the first bioassay tissue was 1 s. After 15 min perfusion to waste, the column effluent was used to superfuse the bioassay tissues.

Figure 2.

Figure 2

Open in a new tab

Diagrammatic representation of the bioassay cascade and the column containing the endothelial cells.

(d) Bioassay

The bioassay was carried out using the cascade superfusion technique (Vane, 1964). Spirally cut strips of bovine coronary artery (BCA; Dusting et al., 1977), rabbit mesenteric or coeliac arteries (RbMA, RbCA; Bunting et al., 1976), rabbit thoracic aorta (RbA; Furchgott, 1955) and occasionally rat stomach strip (RSS; Vane, 1957) were superfused with either warmed (37°C) Krebs solution or with the effluent from the column. The delay between each tissue in the cascade was 3 s. During the experiment, the column could be moved and mounted at any height of the cascade such that its outflow could superfuse any of the assay tissues first. Occasionally a second cascade was superfused in series, with a 4 min delay interposed between them, to study the disappearance of prostacyclin.

(e) Preparation and super fusion of the arterial strips

Fresh bovine hearts were obtained from an abattoir in ice-cold saline. The circumflex and left coronary arteries were dissected 3–4 h after the animals were killed and cut spirally into 4–5 cm long strips which were used immediately for bioassay or stored in ice-cold Krebs buffer for 1–2 days before use. New Zealand White rabbits (2–3 kg) of either sex were anaesthetized with sodium pentobarbitone (40 mg kg−1 i.v.) and RbA, RbMA and RbCA were removed and cut into 3–4 cm long spiral strips.

The arterial strips were mounted in heated (37°C) glass chambers and superfused (5 ml min−1) in cascade with Krebs buffer for 2–3 h before superfusion with effluent from the column. Changes in the length of the tissues were detected by auxotonic levers attached to Harvard 386 heart/smooth muscle transducers, the display of which was recorded by a 6-channel Watanabe recorder.

For assay of EDRF the detector arteries were threaded onto a cotton pipe-cleaner, immersed in Krebs buffer and the endothelium removed by gentle circular and longitudinal movements. The arteries were then cut into spiral strips which were considered to be completely denuded of endothelium when they did not relax to an infusion of the muscarinic agonists 5-methyl-furmethide (1 μm) and acetylcholine bromide (1 μm) or the calcium ionophore A23187 (1 μm). For assay of EDRF the endothelium-denuded arterial strips were contracted with a continuous infusion over the tissues (O.T.) of the 11,9-epoxymethano analogue of prostaglandin H2–U46619 (15–30 Nm), or with phenylephrine hydrochloride (50–200 Nm). Stable stimulators of smooth muscle guanylate cyclase such as sodium nitroprusside (NP) or glyceryl trinitrate (GTN) were infused (10–200 Nm O.T.) to calibrate the relative sensitivity of the RbAs to guanylate cyclase-mediated relaxation.

Stimuli for the release of prostanoids or EDRF from the column were given as one minute infusions. Test substances were infused either O.T. or through the column (T.C.) using a peristaltic minipump (Pierce) or a syringe pump (Braun Perfusor VI). Short infusions (up to 60 s) were given manually using 1 ml polypropylene syringes. Indomethacin (5 μm) was always infused O.T. to inhibit the synthesis of prostanoids by the assay tissues. When inhibition of prostanoid synthesis by the cells in the column was required, the infusion of indomethacin was moved from O.T. to T.C.

When the release of prostanoids or EDRF was attempted with acetylcholine, 5-hydroxytryptamine, noradrenaline or angiotensin II, the direct musculotropic effects of these compounds were blocked by infusing specific antagonists such as atropine sulphate (1.0 μm), phenoxybenzamine (2.0 μm), methysergide (0.3 μm) and saralasine (0.2 μm) O.T.

(f) Radioimmunoassay of prostanoids

The superfusate from some experiments in which prostacyclin and PGE2 were bioassayed was collected, after superfusing the assay tissues, in consecutive fractions of 2.5 ml for radioimmunoassay (RIA) of 6-keto-PGF1α and PGE2 as previously described (Salmon, 1978).

(g) Chemicals

The following chemicals were used: prostacyclin, glyceryl trinitrate, 5-methyl-furmethide, trypsin-versene (Wellcome); sodium nitroprusside, angiotensin II, saralasine, arachidonic acid, atropine sulphate, 5-hydroxytryptamine creatinine sulphate, noradrenaline hydrochloride, formyl-methionyl-leucyl-phenylalanine (FMLP) platelet-activating factor (Paf)-acether, adenosine diphosphate, acetylcholine bromide, Type II collagenase and phenylephrine hydrochloride (Sigma); bradykinin (Sandoz or Sigma); A23187 (Calbiochem); indomethacin (Merck, Sharp and Dohme); methysergide (Sandoz); phenoxybenzamine (Smith, Kline and French); prostaglandin E2 and U46619 (Upjohn Co.); HEPES buffered Dulbecco's modification of Eagle's medium and antibacterial-antimycotic solution (Gibco); foetal calf serum (Flow Laboratories) and Cytodex 3 (Pharmacia).

Results

Release of prostanoids

For bioassay of prostanoids, the cascade comprised BCA, RbMA or RbCA and RSS. These tissues responded to 1 min infusions (O.T.) of prostacyclin and PGE2 in a characteristic fashion (Figure 3). PGE2 was 10 to 20 times more potent than prostacyclin in inducing relaxation of the RbMA and RbCA. The BCA was contracted by PGE2 and relaxed by prostacyclin. The RSS contracted to PGE2 but not to prostacyclin at concentrations up to 300 Nm. In the presence of indomethacin (5 μm, O.T.) the detection limits for both prostanoids was between 10 and 30 Nm.

Figure 3.

Figure 3

Open in a new tab

Effects of bradykinin (Bk, infused over the tissue (O.T.) or through the column (T.C.)), prostaglandin E2 (PGE2) and prostacyclin (PGI2, O.T.) on rabbit mesenteric artery (RbMA) bovine coronary artery (BCA) and rat stomach strip (RSS). Bk (20 Nm, O.T.) induces a contraction of the RSS and has no effect on RbMA and BCA. The same concentration of Bk given T.C. induces the release of a substance(s) which relax(es) the RbMA and the BCA. PGE2 (30 Nm, O.T.) induces a relaxation of the RbMA and a contraction of the BCA and RSS while PGI2 (100 Nm, O.T.) relaxes RbMA and BCA without much effect on the RSS. The effect of mixtures of PGE2 and PGI2 at different concentrations O.T. are also shown.

Bradykinin (Bk; 10–50 Nm, T.C.) or A23187 (0.5–2μm, T.C.) induced, in a concentration-dependent manner, the release of a mixture of prostacyclin and PGE2 as shown by the presence of an unstable BCA relaxing substance (half-life≏4 min) and a stable RbMA/RbCA relaxing material which also contracted the RSS. Calculation of the absolute amounts of prostacyclin and PGE2 released by the column was difficult since both prostanoids had synergistic (RbMA/RbCA) or opposing activities (BCA) on the assay tissues (Figure 3). Nonetheless, bioassay of mixtures of known amounts of prostacyclin and PGE2 enabled the amount of each prostanoid released to be estimated. In nine experiments in which 6 to 11 day old cultures containing 4.4–8.6 × 107 cells per column were used, Bk (20 Nm) T.C. released 9.6 ± 1.5 Nm per 106 cells of prostacyclin and 2.1 ± 0.6 Nm per 106 cells of PGE2 (mean ± s.e.mean). The ratio of the released prostacyclin/PGE2 was 4.6. Because of the difficulties in the bioassay, the amounts of prostanoids released were also measured by specific RIA (Figure 4). Using this method, the amounts of 6-keto-PGF1α and PGE2 released were lower than those detected by bioassay; i.e. 2.9 ± 0.3 Nm per 106 cells for 6-keto-PGF1α and 1.1 ± 0.2 Nm per 106 cells for PGE2 (mean ± s.e.mean, n = 7). The ratio of 6-keto-PGF1α/PGE2 was 2.6

Figure 4.

Figure 4

Open in a new tab

(a) Bioassay of prostacyclin (relaxations of the bovine coronary artery) and (b) radioimmunoassay of 6-keto-prostaglandin F1α (6-keto-PGF1α ○) and PGE2 (▪) released by bradykinin (Bk, 20 Nm), and A23187 (2 μm) infused through the column (T.C.).

In contrast to Bk, A23187 induced a long lasting and non-reproducible release of prostanoids (Figure 4). When calculated on a molar basis A23187 was approximately 100 times less potent than Bk in releasing peak concentrations of prostanoids (Figure 4). AA at concentrations of 10–30μm T.C. produced a moderate release of a BCA-relaxing material that decomposed by approximately 50% after passing through a 4 min delay coil, indicating prostacyclin generation (data not shown).

The following test substances failed to release prostacyclin from the column when administered T.C. at the concentrations given (in μm); acetylcholine bromide 0.5; bethanechol 0.5; noradrenaline hydrochloride 1.0; 5-hydroxytryptamine creatinine sulphate 0.2; adenosine diphosphate 0.25; angiotensin II 0.1; Paf-acether 1.0 and FMLP 1.0. The above concentrations of these substances were the highest that could be used without affecting the tone of the assay tissues when administered O.T.

Release of endothelium-derived relaxing factor

For bioassay of EDRF, the cascade consisted of three or four RbAs denuded of endothelium and submaximally contracted by U46619 (19–30 Nm) or phenylephrine (50–200 Nm). Some of the spiral strips did not respond to U46619 and therefore were contracted with phenylephrine.

In the experiments designed for the simultaneous detection of EDRF and prostacyclin, BCAs denuded of endothelium were included in the cascade (Figure 5). These tissues develop spontaneous tone in the cascade and therefore their response to U46619 or to phenylephrine was smaller than that of the RbAs.

Figure 5.

Figure 5

Open in a new tab

Relaxations of contracted rabbit aorta strips (RbAs) and bovine coronary arteries (BCAs) induced by sodium nitroprusside (NP, 40 Nm, O.T.), prostacyclin (PGI2, 270 Nm, O.T.) and bradykinin (Bk, 100 Nm, T.C.). Bk (100 Nm, O.T.) has no effect on the bioassay tissues. The tissues were contracted by U46619 (30 Nm). Abbreviations as in Figure 3 legend.

EDRF relaxed RbAs and BCAs, whereas prostacyclin relaxed BCAs and had no effect on the tone of RbAs (Figure 5). EDRF was characterized in our bioassay system as a RbA-relaxing substance that could be released from the column by chemical (Bk and A23187; Figure 6) or mechanical stimulation (gentle compression of the cells in the column during the packing of the column or during the experiment, Figure 7). EDRF was very unstable, as shown by the progressive attenuation of the relaxation of the RbAs which were separated from the upper layer of the endothelial cells in the column by delays of 1, 4, 7 and 10 s, respectively. By moving the column down the cascade, it was possible to reduce the delay to each RbA to 1 s (Figure 8). This allowed measurement of the sensitivity of each RbA to EDRF and a clear demonstration of its instability. In most of the experiments the half-life of EDRF appeared to vary between 4 and 7 s, however, detailed calculation of half life was not performed since the rate of disappearance of EDRF seemed to be influenced by the amount detected by the first tissue in the cascade.

Figure 6.

Figure 6

Open in a new tab

Relaxations of contracted rabbit aortic strips (RbAs) induced by bradykinin (Bk, 30 Nm, O.T.), sodium nitroprusside (NP, 200 Nm, O.T.) and Bk (30 and 100 Nm, T.C.). A23187 (300 Nm, O.T.) has very little effect on the assay tissues but induces the release of a labile relaxing substance when given T.C. (300 Nm). The tissues were contracted by U46619 (15 Nm). Abbreviations as in Figure 3 legend.

Figure 7.

Figure 7

Open in a new tab

Relaxation of contracted rabbit aortic strips (RbAs) and bovine coronary artery (BCA) induced by the effluent of a column packed with cells (CE). The effects of prostacyclin (POI2, 100 Nm) and sodium nitroprusside (NP, 100 Nm) are also shown. The tissues were contracted by phenylephrine (50 Nm). Abbreviations as in Figure 3 legend.

Figure 8.

Figure 8

Open in a new tab

Relaxation of contracted rabbit aortic strips (RbAs) induced by sodium nitroprusside (NP, 100 Nm, O.T.). Bradykinin (Bk, 20 Nm) has no effect when given O.T., however, when given T.C. it induces the release of a labile relaxing substance. The second and third Bk infusions T.C. are given when the effluent from the column directly superfuses the second and the third tissues in the cascade. The tissues were precontracted by U46619 (15 Nm). Abbreviations as in Figure 3 legend.

The release of EDRF induced by Bk (10–100 Nm) or A23187 (30–300 Nm) was concentration-dependent and was not inhibited by indomethacin. Bk induced a short-lasting release which tended to disappear during the 1 min infusion whereas the release by A23187 was longer lasting (Figure 6). The release of EDRF was optimal with columns of microcarriers of about 1 ml containing 1–2 × 107 cells, unlike the release of prostanoids which increased in proportion to the number of cells packed in the column. The release of EDRF by Bk was detectable with concentrations between 1 and 3 Nm and was maximal at concentrations between 20 and 100 Nm. The release induced by the first challenge was always greater than that induced by subsequent challenges (Figure 6) which were then reproducible for one or two hours, after which a decline in the release was observed. In contrast, A23187 induced a sustained release which was difficult to reproduce and therefore was not used regularly. EDRF was not released from the column by S-hydroxytryptamine, noradrenaline, Paf-acether, FMLP, acetylcholine or bethanechol at the concentrations described above.

AA at concentrations up to 100 μm had no effect on the bioassay tissues either given O.T. or T.C. At higher concentrations (300 μm O.T.) it relaxed the assay tissues. When administered T.C. it produced an additional increase in relaxation in the first RbA of the cascade.

Discussion

Vascular endothelium generates two unstable vasodilators, prostacyclin (Moncada et al., 1976; Gryglewski et al., 1976) and EDRF (Furchgott & Zawadzki, 1980). Because of the lack of stability of prostacyclin and of EDRF in their physiological environment (half-lives of about 3 min and 6 s respectively) they can only be detected by biological methods.

The elucidation of the structure of prostacyclin and its chemical degradation product, 6-keto-PGF1α, has permitted the use of other methods like RIA, high pressure liquid chromatography and gas chromatography/mass spectrometry for the detection and quantitation of prostacyclin. At present, however, our knowledge of EDRF is very limited and bioassay is the only technique by which it may be studied. Until now, the evidence for the release of EDRF and its action on tissues away from the site of release comes from experiments in which the products of arteries with intact endothelium are detected on arteries denuded of endothelium (Furchgott, 1984; Griffith et al., 1984). More recently Cocks et al. (1985) have described experiments similar to our own in which the effluent of a volume of endothelialised beads packed in the barrel of a syringe was used to superfuse a single arterial ring.

We have now adapted the bioassay superfusion cascade of Vane (1964) to the assay of EDRF. This method has been successfully used to study the release of many vasoactive substances from a variety of perfused organs (Vane, 1969; Moncada et al., 1978). In the present study we have replaced the heterogeneous population of cells which are present in any perfused organ with a homogeneous population of cultured vascular endothelial cells on microcarrier beads, packed in a chromatographic column and perfused in the same way as an isolated organ. The effluent from this column superfused in a cascade a series of vascular smooth muscle strips which detect prostacyclin and PGE2 such as BCA (Dusting et al., 1977), RbMA/RbCA (Bunting et al., 1976) and RSS (Vane, 1957). Inclusion in the cascade of RbA strips (Furchgott, 1955) contracted and denuded of endothelium (Furchgott & Zawadski, 1980) permitted the simultaneous bioassay of EDRF (see Figure 5).

Cultures of porcine aortic endothelial cells (6–14 days old) do not release prostanoids or EDRF in response to stimulation by some receptor agonists at the concentrations used in this study. Other authors have, however, demonstrated release of prostacyclin with higher concentrations of adenosine diphosphate and angiotensin II under similar experimental conditions (Pearson et al., 1983b). Our results indicate that these cells lose their responsiveness to these agonists after a few days in culture, confirming a previous finding (Cocks et al., 1985). However, these agonists do release prostacyclin and/or EDRF from intact vascular tissue in vitro and in vivo (Furchgott & Zawadski, 1980; Beetens et al., 1982; Gryglewski, 1982; Cocks & Angus, 1983; Forstermann & Neufang, 1984). Paf-acether did not release prostacyclin or EDRF from the column.

Bk and A23187 were the most potent releasers of prostacyclin and EDRF, therefore Bk (20–100 Nm) was used as the standard stimulus in most of our studies. The release of EDRF induced by both Bk and A23187 is temporally similar to the release of prostanoids induced by infusions of these agents into guinea-pig isolated lungs. Release induced by Bk is rapid, short-lasting and disappears during the infusion (Palmer et al., 1973; Bakhle et al., 1985) while that induced by A23187 is slow, long-lasting and reaches a peak several minutes after the end of the infusion (Bakhle et al., 1985). A23187 is generally considered to be the most potent releaser of prostacyclin and EDRF from vascular endothelium (Peach et al., 1985). However, in our system Bk is considerably more potent than A23187 in releasing prostacyclin and marginally more potent in releasing EDRF, confirming the findings of Cocks et al. (1985).

It is interesting that unlike prostacyclin the release of EDRF does not increase with the number of cells in the column. It is likely that this is due to inactivation of EDRF as it passes from the lower to the upper layers of endothelium inside the column.

AA (10–30 μm) released prostacyclin but not EDRF. Ten fold higher concentrations were required to generate a trace of EDRF. Unlike cyclo-oxygenase, simple additions of substrate are not enough for the lipoxygenase to generate its products (Borgeat & Samuelsson, 1979). Therefore the small and erratic release of EDRF following stimulation with high concentrations of AA neither confirmed nor excluded the possibility that EDRF might be a metabolite of this fatty acid (Furchgott & Zawadzki, 1980; Forstermann & Neufang, 1984; Peach et al., 1985).

Both prostacyclin and EDRF were also released from the column by mechanical stimulation (Figure 7). Indeed, this may be a good method of producing large amounts of EDRF or its degradation products.

Bk and A23187-induced release of prostacyclin was recorded as the relaxation of BCA and RbMA/RbCA. However, when compared with the standard prostacyclin response, the Bk-induced release of prostacyclin from endothelial cells was obviously influenced by the presence of another substance which contracted BCA and relaxed RbMA/RbCA more profoundly than prostacyclin itself (Figure 3). This substance was identified as PGE2 from its profile of biological activity (Figure 3) and RIA data (Figure 4). Concomitant release of prostacyclin and PGE2 occurred at a ratio 2.6–4.6 compromising the precise quantification of those two prostanoids by bioassay. The bioassay of prostacyclin, therefore, poses more problems than that of EDRF. The interference of EDRF with the bioassay of prostacyclin is easily eliminated by moving BCAs down the cascade to create a delay of at least 10 s from the upper layer of endothelial cells in the column.

The bioassay of EDRF is not influenced by the presence of prostacyclin or PGE2 since neither of these two prostanoids affected the tone of the RbAs. We have shown that inhibition of prostanoid generation by indomethacin does not influence the generation of EDRF by endothelial cells as previously reported (Furchgott et al., 1984; Peach et al., 1985), thus EDRF can be assayed on its own if the cells in the column are treated with indomethacin.

Bioassay of EDRF alone requires 3–4 RbAs mounted in a cascade and separated by a 3 s delay between each. The attenuation of RbA relaxations down the cascade is characteristic of EDRF. In most of the experiments EDRF did not survive for 7 s; indeed, in some experiments a 4 s delay was sufficient for EDRF to disappear (Figure 7) indicating that the half-life of EDRF was, on average, less than 7 s. This disappearance of EDRF down the cascade did not seem to be linear but rather depended on the amount of EDRF being released, as recorded by the relaxation of the uppermost tissue in the cascade (Figure 6). Detailed analysis of this phenomenon is difficult without a knowledge of the dose-response curve of EDRF on the bioassay tissues.

In conclusion, we have described a method for the bioassay of prostacyclin and EDRF resulting from the combination of several established techniques. The disadvantages of this method include the following: firstly, the generation of the endothelium-derived vasodilators occurs in cultured endothelial cells which have probably lost their receptors for several physiological agonists. Secondly the cells are separated from the underlying smooth muscle so a biochemical collaboration between those two vascular layers cannot be studied. Finally, the bioassay of prostacyclin is influenced by the presence of PGE2 in the effluent. On the other hand, the advantage of this technique include the simultaneous detection of prostacyclin and EDRF released from endothelial cells; the lack of interference of these vasodilators on bioassay tissues and the potential for studying the effects of chemical and physical factors on the generation of the endothelium-derived vasodilators.

Acknowledgments

We thank Ms L. Pynegar and Mr N.A. Foxwell for technical assistance, Dr G. De Nucci for carrying out the radioimuno-assay and Mr P.J. Astbury for the electron micrography. We also thank Sir John Vane for his helpful discussion.

References

  1. Ager A, Gordon JL, Moncada S, Pearson JD, Salmon JA, Trevethic MA. Effects of isolation and culture on prostaglandin synthesis by porcine aortic endothelial and smooth muscle cells. J. cell Physiol. 1982;110:9–16. doi: 10.1002/jcp.1041100103. [DOI] [PubMed] [Google Scholar]
  2. Bakhle YS, Moncada S, Nucci G, De Salmon JA. Differential release of eicosanoids by bradykinin, arachidonic acid and calcium ionophore A23187 in guinea-pig isolated perfused lung. Br. J. Pharmac. 1985;86:55–62. doi: 10.1111/j.1476-5381.1985.tb09434.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Beetens JR, Van Hove C, Rampart M, Herman AG. Acetylcholine stimulates the release of prostacyclin by rabbit aorta endothelium. J. Pharm. Pharmac. 1982;35:251–252. doi: 10.1111/j.2042-7158.1983.tb02924.x. [DOI] [PubMed] [Google Scholar]
  4. Borgeat P, Samuelsson B. Arachidonic acid metabolism in polymorphonuclear leukocytes: effects of ionophore A23187. Proc. nam. Acad. Sci., U.S.A. 1979;76:2148–2152. doi: 10.1073/pnas.76.5.2148. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bunting S, Moncada S, Vane JR. The effects of prostaglandin endoperoxides and thromboxane A2 on strips of rabbit coeliac artery and certain other smooth muscle preparations. Br. J. Pharmac. 1976;57:462P–463P. [PMC free article] [PubMed] [Google Scholar]
  6. Cocks TM, Angus JA. Endothelium-dependent relaxation of coronary arteries by noradrenaline and serotonin. Nature. 1983;305:627–630. doi: 10.1038/305627a0. [DOI] [PubMed] [Google Scholar]
  7. Cocks TM, Angus JA, Campbell JH, Campbell GR. Release and properties of endothelium-derived relaxing factor (EDRF) from endothelial cells in culture. J. cell Physiol. 1985;123:310–320. doi: 10.1002/jcp.1041230304. [DOI] [PubMed] [Google Scholar]
  8. Davies PF. Microcarrier culture of vascular endothelial cells on solid plastic beads. Exp. cell Res. 1981;134:367–376. doi: 10.1016/0014-4827(81)90436-5. [DOI] [PubMed] [Google Scholar]
  9. Dusting GJ, Moncada S, Vane JR. Prostacyclin (PGX) is the endogenous metabolite responsible for relaxation of coronary arteries induced by arachidonic acid. Prostaglandins. 1977;13:3–15. doi: 10.1016/0090-6980(77)90037-5. [DOI] [PubMed] [Google Scholar]
  10. Forstermann U, Neufang B. The endothelium-dependent vasodilator effect of acetylcholine: characterization of the endothelial relaxing factor with inhibitors of arachidonic acid metabolism. Eur. J. Pharmac. 1984;103:65–70. doi: 10.1016/0014-2999(84)90190-0. [DOI] [PubMed] [Google Scholar]
  11. Furchgott RF. Pharmacology of vascular smooth muscle. Pharmac. Rev. 1955;7:183–265. [PubMed] [Google Scholar]
  12. Furchgott RF. The role of endothelium in the responses of vascular smooth muscle to drugs. A Rev. Pharmac. Tox. 1984;24:175–197. doi: 10.1146/annurev.pa.24.040184.001135. [DOI] [PubMed] [Google Scholar]
  13. Furchgott RF, Zawadzki JV. The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature. 1980;288:373–376. doi: 10.1038/288373a0. [DOI] [PubMed] [Google Scholar]
  14. Furchgott RF, Cherry PD, Zawadzki JV, Jothianandan D. Endothelial cells as mediators of vasodilatation of arteries. J. cardiovasc. Pharmac. 1984;6:S336–S343. doi: 10.1097/00005344-198406002-00008. [DOI] [PubMed] [Google Scholar]
  15. Gimbrone MA, Jr., Alexander RW. Angiotensin II stimulation of prostaglandin production in cultured human vascular endothelium. Science. 1975;189:219–220. doi: 10.1126/science.1138377. [DOI] [PubMed] [Google Scholar]
  16. Gordon JL, Martin W. Stimulation of endothelial prostacyclin production plays no role in endothelium-dependent relaxation of the pig aorta. Br. J. Pharmac. 1983;80:179–186. doi: 10.1111/j.1476-5381.1983.tb11064.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Griffith TM, Edwards DH, Lewis MJ, Newby AC, Henderson AH. The nature of endothelium-derived vascular relaxant factor. Nature. 1984;308:645–647. doi: 10.1038/308645a0. [DOI] [PubMed] [Google Scholar]
  18. Gryglewski RJ. Regulation of prostacyclin release into circulation. In: Yoshida H, Hagihara V, Ebashi S, editors. Advances in Pharmacology and Therapeutics. Vol. 4. Oxford: Pergamon Press; 1982. pp. 235–249. [Google Scholar]
  19. Gryglewski RJ, Bunting S, Moncada S, Flower RJ, Vane JR. Arterial walls are protected against deposition of platelet thrombi by a substance (Prostaglandin X) which they make from prostaglandin endoperoxides. Prostaglandins. 1976;12:685–714. doi: 10.1016/0090-6980(76)90047-2. [DOI] [PubMed] [Google Scholar]
  20. Hong SL. Effect of bradykinin and thrombin on prostaglandin synthesis in endothelial cells from calf and pig aorta and human umbilical cord vein. Thromb. Res. 1980;18:787–795. doi: 10.1016/0049-3848(80)90201-7. [DOI] [PubMed] [Google Scholar]
  21. Johnson AR. Human pulmonary endothelial cells in culture. Activities of cells from arteries and cells from veins. J. clin. Invest. 1980;65:841–850. doi: 10.1172/JCI109736. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Malick LEA, Wilson RB. Modified thiocarbohydrazide procedure for scanning election microscopy: routine use for normal, pathological or experimental tissues. J. Stain Technol. 1975;50:265–269. doi: 10.3109/10520297509117069. [DOI] [PubMed] [Google Scholar]
  23. Moncada S, Ferreira SH, Vane JR. Bioassay of prostaglandins and biologically active substances derived from arachidonic acid. In: Frolich JC, editor. Advances in Prostaglandin and Thromboxane Research. Vol. 5. New York: Raven Press; 1978. pp. 211–236. [PubMed] [Google Scholar]
  24. Moncada S, Gryglewski RJ, Bunting S, Vane JR. An enzyme isolated from arteries transforms prostaglandin endoperoxides to an unstable substance that inhibits platelet aggregation. Nature. 1976;263:663–665. doi: 10.1038/263663a0. [DOI] [PubMed] [Google Scholar]
  25. Palmer MA, Piper PJ, Vane JR. Release of rabbit aorta contracting substance (RCS) and prostaglandins induced by chemical or mechanical stimulation of guinea-pig lungs. Br. J. Pharmac. 1973;49:226–242. doi: 10.1111/j.1476-5381.1973.tb08368.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Peach MJ, Singer HA, Loeb AL. Mechanisms of endothelium-dependent vascular smooth muscle relaxation. Biochem. Pharmac. 1985;34:1867–1874. doi: 10.1016/0006-2952(85)90300-4. [DOI] [PubMed] [Google Scholar]
  27. Pearson JD, Carleton JS, Hutchings A. Prostacyclin release in response to thrombin or bradykinin by cultured porcine endothelial cells derived from aorta or umbilical vein. Thromb. Res. 1983a;29:115–124. doi: 10.1016/0049-3848(83)90133-0. [DOI] [PubMed] [Google Scholar]
  28. Pearson JD, Slakey LL, Gordon JL. Stimulation of prostaglandin production through purinoceptors on cultured porcine endothelial cells. Biochem. J. 1983b;214:273–276. doi: 10.1042/bj2140273. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Ryan US, Mortara M, Whitaker C. Methods for microcarrier culture of bovine pulmonary artery endothelial cells avoiding the use of enzymes. Tissue Cell. 1980;12:619–635. doi: 10.1016/0040-8166(80)90017-8. [DOI] [PubMed] [Google Scholar]
  30. Salmon JA. A radioimmunoassay for 6-keto-prostaglandin F1α. Prostaglandins. 1978;15:383–397. doi: 10.1016/0090-6980(78)90122-3. [DOI] [PubMed] [Google Scholar]
  31. Sanford KK, Earle WR, Evans VJ. The measurement of proliferation in tissue cultures by enumeration of cell nuclei. J. Nat. Cancer Inst. 1951;11:773–795. [PubMed] [Google Scholar]
  32. Vane JR. A sensitive method for the assay of 5-hydroxytryptamine. Br. J. Pharmac. Chemother. 1957;12:344–349. doi: 10.1111/j.1476-5381.1957.tb00146.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Vane JR. The use of isolated organs for detecting active substances in the circulating blood. Br. J. Pharmac. Chemother. 1964;23:360–373. doi: 10.1111/j.1476-5381.1964.tb01592.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Vane JR. The release and fate of vasoactive hormones in the circulation. Br. J. Pharmac. 1969;35:209–242. doi: 10.1111/j.1476-5381.1969.tb07982.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Weksler BB, Ley CW, Jaffe EA. Stimulation of endothelial cell prostacyclin production by thrombin, trypsin and the ionophore A23187. J. clin. Invest. 1978;62:923–930. doi: 10.1172/JCI109220. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from British Journal of Pharmacology are provided here courtesy of The British Pharmacological Society

RESOURCES