A Live-Cell Assay for Studying Extracellular and Intracellular Endothelin-Converting Enzyme Activity
Abstract Endothelin-1 (ET-1) is formed from its precursor preproET-1 via the cleavage of the intermediate bigET-1 by endothelin-converting enzyme (ECE-1). However, the subcellular site at which this step occurs is not clear: It could occur intravesicularly along the secretory pathway or bigET-1 might be released and processed extracellularly. To address this point, we have developed an integrated autocrine system that uses a recombinant Chinese hamster ovary (CHO) luciferase reporter cell line that permanently expresses the human ETA receptor. Into these cells we transiently transfected human ECE-1a cDNA, either together with the human preproET-1 cDNA (as an endogenous source of bigET-1), or alone (in which case exogenous bigET-1 was added). Phosphoramidon inhibited the conversion of exogenous bigET-1 (IC50=5 to 30 μmol/L) much better than that of endogenous bigET-1 (IC50 >1 mmol/L). Both conversions showed similar high yields (20% to 100%) that depended on the amount of ECE-1a expressed. Thus, ECE-1a has two equally relevant activities in this recombinant system for CHO cells: (1) an intracellular, probably intravesicular activity, corresponding to the ECE-1a–mediated step of ET-1 biosynthesis and (2) an extracellular activity at the plasma membrane. If this is also the case for endothelial cells, ECE-1a inhibitors would have to cross the plasma and vesicle membranes to be effective. The present system could be useful for screening such inhibitors.
The endothelins are a recently discovered family of three 21–amino acid peptides that may regulate vascular tone and blood pressure.1 2 Endothelin-1 (ET-1) was first isolated from endothelial cells due to its high vasoconstrictive activity,3 and is derived from a longer precursor molecule preproET-1 (212 amino acids). The two other isotypes, endothelin-2 (ET-2) and endothelin-3 (ET-3), were identified a few months later. ET-2 and ET-3 differ from ET-1 by 2 and 6 amino acids, respectively, and as for ET-1 are produced from precursor molecules (preproET-2 and preproET-3, 178 and 238 amino acids, respectively) that are encoded by separate genes.4 The preproETs are first cleaved at two pairs of basic amino acids5 to generate bigET-1, bigET-2, and bigET-3 (38 to 41 amino acids). These peptides are then cleaved by endothelin-converting enzyme (ECE) (EC. 3. 4. 24. 71) at an unusual Trp-Val (for bigET-1 and bigET-2) or Trp-Ile (for bigET-3) site to produce the 21-residue active peptides.3 6
ECE is a key enzyme in the biosynthesis of the endothelins because the biological activities of bigETs are negligible.7 8 Its identification was particularly difficult, and various enzyme activities were proposed to be the physiological ECE. However, there is now good evidence that ECE is a new membrane-bound phosphoramidon-sensitive metalloprotease.6 9 10 Indeed, the ECE in endothelial cells is inhibited by phosphoramidon11 12 13 but not by thiorphan, a specific inhibitor of another metalloprotease, neutral endopeptidase (NEP).14 The ECE-1 isoforms have been purified from bovine adrenal cortex15 and from rat lung16 and cloned from cDNA libraries of the corresponding endothelial cells. Two human ECE-1 isoforms, ECE-1a and ECE-1b,17 18 19 have recently been cloned that differ only by their N-terminal extremities and are encoded by a single gene.20
The primary structure of ECE-1 indicates that it is indeed a member of the zinc metalloprotease family, harboring the HEXXH motif, and displaying 40% similarity to NEP. ECE-1 is assumed to be a type II integral membrane protein with a short N-terminal cytoplasmic tail, a single transmembrane hydrophobic helix, and a large putative extracellular domain containing the catalytic site. Membrane preparations of COS cells transfected with the cloned ECE-1 produced a good conversion of bigET-1 in vitro and a less efficient conversion of bigET-2 and bigET-3, suggesting that this isoform is specifically implicated in ET-1 synthesis.15 16 Since endothelial cells express both preproET-1 and ECE-1, ECE-1 is likely to be the enzyme responsible for the production of ET-1 in vivo. Therefore, ECE-1 is a new pharmaceutical target and ECE-1 inhibitors may be useful in the treatment of various disorders linked to ET-1 levels. Therefore, a clear understanding of how ECE-1 functions is of primary importance.
One major question is the exact site of ECE-1–mediated conversion of endogenous bigET-1 to ET-1. In the secretory pathway, the ECE-1 catalytic site faces the interior of secretion vesicles and then the exterior of the cell. Hence, the ECE-1–mediated step could occur within vesicles or endogenous bigET-1 could be released outside the cell and subsequently processed at the cell surface. If the conversion of bigET-1 occurs within the vesicle, then the function of ECE-1 at the cell surface is unclear. The results of immunohistochemical studies on cultured endothelial cells are contradictory, showing that ECE-1 forms large clusters at the plasma membrane21 and that ET-1 is localized within intracellular vesicles.22 The conclusions from these two observations are contradictory, one indicating a processing at the plasma membrane and the other within intracellular vesicles. The location of recombinant ECE-1 expressed in CHO cells has also been examined by Xu et al.15 This study reported that the conversion of endogenous bigET-1 (supplied by the transient transfection of preproET-1 cDNA) was more efficient than the conversion of exogenous bigET-1 (supplied by coculture of a CHO cell line permanently expressing preproET-1) and that the inhibitor FR901533 blocked the conversion of exogenous bigET-1 but not that of the endogenous precursor. The conclusion drawn from this study was that the activity of ECE-1 is mainly intracellular. However, Takahashi et al21 observed clusters of ECE-1 in CHO transfected cells exactly as in endothelial cells, and suggested that ET-1 is processed at the plasma membrane.
An intravesicular production of ET-1 by ECE-1 would have important consequences. First, it would be an unusual example of intracellular activity by a member of the NEP family. Second, it would imply that ECE-1 inhibitors have to cross the plasma membrane and the vesicle membrane to inhibit ET-1 production in endothelial cells in vivo, indicating that in vitro assays are not suitable for screening potential inhibitors.
To investigate further the site of bigET-1 conversion to ET-1, we have developed an autocrine-integrated cell system. The binding of ET-1 to its receptor stimulates the AP-1 transcription factor,23 which binds to the TRE to activate the transcription of genes carrying this sequence in their promoter. We established a recombinant CHO cell line permanently expressing both the hETA receptor and a luciferase reporter gene construct carrying TRE sequences in its promoter. The human ECE-1a cDNA was then transiently transfected into these cells, either together with the human preproET-1 cDNA (as an endogenous source of bigET-1) or alone (in which case bigET-1 was added exogenously), allowing the study and comparison of the conversion of endogenous and exogenous bigET-1s (Fig 1⇓). This cell line provides a novel original live-cell assay for studying the activity of ECE-1a.
The cDNA coding for human preproET-1 was generated by reverse transcription–polymerase chain reaction using mRNA from primary cultures of human umbilical vein endothelial cells. The primers, chosen from the published sequence,24 were 5′-GCTAGCCCGGGATGGATTATTTGCTCATGATTTTC-3′ (sense primer) and 5′-GTCGACCTCGAGTCACCAATGTGCTCGGTTGTGGGT-3′ (reverse primer). The complete cDNA was subcloned into the Sma I-Xho I site (underlined) of the pece expression vector.25
The cDNA coding for the human ETA receptor was generated by reverse transcription–polymerase chain reaction using human placenta mRNA (kindly provided by Dr S. Nadaud). The primers were chosen from the sequence published by Hosada et al,26 and were 5′-CTTCTCGAGATCGATATG GAAACCCTTTGCCTCAGGGCA-3′ (sense primer) and 5′-CTTCCGCGGAAGCTTTCAGTTCATGC TGTCCTTATG GCT GCT-3′ (reverse primer). The underlined nucleotides are restriction enzyme sites, Xho I followed by Cla I, and HindIII followed by Sac II, respectively. The complete cDNA was subcloned into the pcDNA3 expression vector (InVitrogen).
The cDNA coding for human ECE-1a was generated by reverse transcription–polymerase chain reaction using mRNA from primary cultures of human umbilical vein endothelial cells in two steps. First, a partial cDNA corresponding to the nucleotide sequence 1 to 1770 was generated using the sense primer 5′-CCGTGGGAACCAGACCACCCCTGA-3′ and the reverse primer 5′-AGGCGTTCACCATGGGCGGGGTCA-3′, then a partial cDNA corresponding to the nucleotide sequence 1220 to 2395 was generated using the sense primer 5′-ACGGAAGACGAGCTCCTTCCTCGA-3′ and the reverse primer 5′-CAGCTCGAGGATCCTCACCAGACTTCACACTTGTGATG-3′. The underlined nucleotides are restriction enzyme sites, Sac I and BamHI. Both of the polymerase chain reaction generated cDNAs were digested by BamHI (the second was only partially digested) and the full cDNA coding for ECE-1a was subcloned into the pcDNA3 expression vector.
Serial NxTRE-Coll-LUC reporter plasmids were obtained by constructing a 0xTRE-Coll-LUC plasmid by insertion into the Sac I–Bgl II sites of pGL2 Basic Vector (Promega) containing the luciferase reporter gene of the following double-stranded oligonucleotide 5′-GAGCTCACTGTGTCGACGCGTGCAAG GACTCTATATATACAGAGGGAGCTTCCTAGCTG GGATATTGGAGCAGCAAGAGGCTGGGAAGCCATCA CTTACCTTGCACTGAGATCT-3′ (TATA box is in italics and Sac I and Bgl II sites are underlined) coding for the minimal collagenase promoter.27 The double-stranded oligonucleotides 5′-GAGCTCATAAAGCATGAGTCAGACACCTC TGGCTTTCTACGCGT-3′ and 5′-GAGCTCATAAAGCA TGAGTCAGACACCTCTGGCTTTCTGGAAATGAG TCA GACACCTCTGGCTTTCTACGCGT-3′ (TRE sequences are in italics and Sac I and Mlu I sites are underlined) were then inserted into the Sac I–Mlu I site of 0xTRE-Coll-LUC to create 1xTRE-Coll-LUC and 2xTRE-Coll-LUC plasmids; NxTRE-Coll-LUC plasmids (n=3, 4, 5, 6, 7, 8, 10, and 12) were finally generated by iterative insertions of the double-stranded oligonucleotides 5′-CCCGGGTCTAGAGGTACCGAGCT CATAAAGCATGAGTCAGACACCTCTAGCTC-3′ into Sma I–Sac I sites of the successive constructions. The procedure can be used iteratively, because each insertion creates a new Sac I site and destroys the former (GAGCTC sequence replaced by TAGCTC).
ET-1, bigET-1, and SB20967028 were synthesized at Roussel-Uclaf. Cell culture products were purchased from GIBCO BRL, except for Transfectam (Promega). All other chemicals were from Sigma Chemical Co.
Cell Culture and Stable Transfection
CHO cells were grown in Dulbecco’s modified Eagle’s medium/Ham’s F-12 (DMEM/F-12) supplemented with 100 UI/mL penicillin, 100 μg/mL streptomycin, and 7.5% fetal calf serum under a humidified atmosphere of 5% CO2/95% air at 37°C. The day before transfection, a 75-cm2 flask was seeded with 9×105 cells. DNA transfections were performed by calcium-phosphate precipitation29 of 10 μg of each plasmid of interest, pcDNA3-hETA and 6xTRE-Coll-LUC, and 2 μg of each plasmid of selection, pSV-neo and pSG5-puro. The cells were incubated for 4 hours and subjected to a 15% (wt/vol) glycerol shock for 2 minutes. The medium was changed and the cells left for 3 days. They were then incubated in selective medium (750 μg/mL geneticin and 5 μg/mL puromycin) for 2 weeks, with medium changed twice a week. Isolated colonies (about 50) were cultured independently in medium containing 400 μg/mL geneticin and 5 μg/mL puromycin and assayed for the ETA receptor.
ET-1 Binding Assays
Binding assays were performed on intact adherent cells. Cells (4×104 per well) were seeded in white opaque 96-well plates (CulturePlate, Packard) and incubated for 24 hours. Cells were then washed twice with PBS and incubated for 45 minutes at 37°C in HEPES-buffered DMEM/F-12 supplemented with 0.2% lysozyme, 70 pmol/L 125I-labeled ET-1, and various concentrations of unlabeled ET-1, SB209670, or bigET-1. Cells were then washed three times with 200 μL/well of washing buffer (50 mmol/L Tris, 100 mmol/L MgCl2, pH 7.4). Scintillation liquid (Microscint 20, Packard) was then added directly to the cells, and the plates were counted on a TopCount radioactivity counter (Packard).
Cells were seeded and incubated as for the ET-1 binding assays. The culture medium was then changed to 100 μL/well serum-free DMEM/F-12 containing 2 μCi/mL [3H]inositol (19 Ci/mol, Amersham) for 24 hours. The cells were rinsed and incubated in HEPES-buffered DMEM supplemented with 0.2% lysozyme, 13 mmol/L LiCl, and various concentrations of ET-1, SB209670, or bigET-1 for 15 minutes and lysed in 10 mmol/L formic acid. The lysates were analyzed on Dowex AG1×8 anion exchange columns (BioRad) sequentially eluted with 3 mmol/L NH4OH, 40 mmol/L ammonium formate, and 2 mol/L ammonium formate. The last fraction, containing the total inositol phosphates, was counted.
Cells were seeded and incubated as for the ET-1 binding assays, then rinsed, and incubated for 24 hours in serum-free DMEM/F-12 supplemented with 0.2% lysozyme, 1% DMSO, and various concentrations of ET-1, SB209670, or bigET-1. Cells were then rinsed with 200 μL PBS and 50 μL PBS containing Mg2+, and 50 μL LucLite Reagent (Packard) was added to each well. The plates were gently shaken for 5 minutes, and luminescence was measured on a TopCount (Packard) in single photon counting mode.
Cells (2×104 per well) were seeded into white opaque 96-well plates and incubated for 24 hours. One hour before transfection, cells were rinsed and incubated in 100 μL serum-free DMEM/F-12. Transient transfections were performed using Transfectam reagent (Promega). DNA solutions and Transfectam solutions were prepared separately in serum-free DMEM/F-12: The amounts used were 0.5 μg DNA in 50 μL medium and 2 μL Transfectam reagent in 50 μL medium per well. The Transfectam and DNA solutions were mixed and 100 μL of the mixture was added to the cells in each well to give a final volume of 200 μL/well, and the plates were incubated for 24 hours. Transfection was stopped by washing the cells with PBS and adding 100 μL/well DMEM/F-12 supplemented with 7.5% fetal calf serum. The cells were incubated for a further 18 to 24 hours, rinsed with 200 μL serum-free DMEM/F-12 and incubated for 18 to 24 hours in 100 μL serum-free DMEM/F-12 supplemented with 0.2% lysozyme, 1% DMSO, and various concentrations of ET-1, SB209670, bigET-1, phosphoramidon, and thiorphan, at 10% CO2. Finally, luciferase activity was measured.
For normalization with β-Gal activity, independent wells were incubated for the same time with the transfection mixture of interest, rinsed with 200 μL PBS, and incubated for 48 hours in DMEM/F-12 supplemented with 7.5% fetal calf serum. Cells were then lysed with 30 μL lysis buffer (250 mmol/L Tris, 5 mmol/L DTT, 15% glycerol, pH 7.8) and the plates were shaken for 20 minutes and centrifuged for 10 minutes at 1500 rpm. Aliquots of lysate (20 μL) were transferred to a transparent 96-well plate, reaction mixture (52.5 mmol/L Na2HPO4, 35 mmol/L NaH2PO4, 8.75 mmol/L KCl, 875 μmol/L β-mercaptoethanol, 6.25 mmol/L chlorophenolred-β-D-galactopyranoside) was added at 80 μL per well, and the plates were incubated for 1 hour at 37°C. Optical density at 580 nm was measured on a CERES 900 reader (Bio-Tek).
Establishment of a Recombinant CHO Reporter Cell Line for Detecting Human ETA Receptor Activation
A recombinant CHO reporter cell line was established by stable transfection of the cDNA coding for the human ETA receptor (hETA) and the 6xTRE-Coll-LUC plasmid. The effects of ET-1, bigET-1, and SB209670 (ETA,B antagonist) on this cell line were tested for binding to intact cells (Fig 2A⇓), for IP production (Fig 2B⇓), and for luciferase activity (Fig 2C⇓). In binding assays, IC50 was 3 nmol/L for ET-1 and SB209670, and there was no competition with bigET-1 (IC50 >1 μmol/L). In IP assays, ET-1 produced a dose-dependent stimulation, with an EC50 of 4 nmol/L (compatible with data published previously30 ) and a 5-fold maximal stimulation, shifted 1000-fold to the right by 5 μmol/L SB209670. BigET-1 had no effect on IP production (EC50 >1 μmol/L). In luciferase assays, ET-1 induced a dose-dependent stimulation, with an EC50 of 1 nmol/L and a 32-fold maximal stimulation, shifted 500-fold to the right by 5 μmol/L SB209670. The presence of SB209670 also resulted in a decrease of the maximal stimulation factor, which was also observed with other antagonists (data not shown). BigET-1 had a marginal effect on luciferase activity when compared with the effect of ET-1, with a significant 2-fold stimulation (P<.001) from 100 nmol/L bigET-1.
The recombinant CHO reporter cell line was used to perform transient transfections with preproET-1 and ECE-1a. For each transfection, luciferase activity was normalized using exogenous 10 nmol/L ET-1 stimulation, which defines the maximal stimulation (100%) allowed by ET-1 and a mock transfection to define the background stimulation (0%) (Fig 3A⇓). All the luciferase activities were thus expressed as a percentage of exogenous 10 nmol/L ET-1 stimulation over background. The transfection step reduced the amplitude of the response of the cells to ET-1 (8-fold stimulation) as compared with untransfected cells (32-fold stimulation), with no change in the EC50 (Fig 3B⇓). This was not the case for 20 nmol/L bigET-1 applied alone, which therefore had a greater relative effect on luciferase production (25% to 30% of maximal stimulation) under these experimental conditions (Fig 3B⇓). However, transfection of ECE-1a alone or preproET-1 alone induced no luciferase production (Fig 3A⇓). Exogenous bigET-1 induced a luciferase production by ECE-1a–transfected cells comparable to 10 nmol/L ET-1 stimulation, whereas the induction was only 30% of the maximal stimulation in mock or preproET-1 transfected cells (Fig 3A⇓). The double transfection (preproET-1 and ECE-1a) resulted in the same amount of luciferase activity as did the addition of exogenous 20 nmol/L bigET-1 or 10 nmol/L ET-1 (Fig 3A⇓). The induction of luciferase activity was completely blocked by 5 μmol/L SB209670 in all cases (data not shown), showing that luciferase was specifically induced via the hETA receptor.
Effects of Inhibitors on ECE-1a Conversion Activity
The effects of thiorphan, a specific NEP inhibitor,14 and phosphoramidon, an ECE inhibitor,31 were tested on the conversion of exogenous and endogenous bigET-1s. The IC50 for thiorphan was greater than 100 μmol/L for both exogenous and endogenous conversions, which is compatible with the known pharmacology of ECE-1 and shows that the conversion is not due to NEP. Phosphoramidon had differential effects on the two types of conversion (Fig 4⇓). The conversion of exogenous bigET-1 was inhibited in a dose-dependent manner, with an IC50 of 5 μmol/L for 5 nmol/L bigET-1 and 30 μmol/L for 20 nmol/L bigET-1. Conversely, 1 mmol/L phosphoramidon was not sufficient to inhibit the conversion of endogenous bigET-1 (IC50 >1 mmol/L). The lack of effect of phosphoramidon on the conversion of endogenous bigET-1 was not due to the production of ET-1 before adding the inhibitor because similar results were obtained with cells preincubated for 24 hours with 1 mmol/L phosphoramidon. Also, the luciferase activity of mock-transfected cells stimulated with exogenous 10 nmol/L ET-1 for 24 hours and then rinsed, decreased back to the basal level within 10 to 20 hours, indicating that the cessation of ET-1 production should result in lowered luciferase levels even if cells have been stimulated before.
Further Characterization of the Conversion of Exogenous bigET-1
Various amounts of ECE-1a plasmid were transfected, and for each amount, various concentrations of bigET-1 were added to the medium. Fig 5⇓ shows that ECE-1a was not the limiting factor of the assay, because luciferase activity could be detected for as little as 0.1 ng/well ECE-1a plasmid, and high luciferase activity occurred for 1.5 to 50 ng/well ECE-1a plasmid. Various concentrations (1.25 to 20 nmol/L) of exogenous bigET-1 were tested. For maximal luciferase activity, 20 nmol/L bigET-1 was required, showing that the substrate bigET-1 was the limiting factor.
In ECE-1–transfected cells, the exogenous conversion of a given concentration of bigET-1 was deduced from the concentration of exogenous ET-1 that induced the same luciferase activity. For example, cells transfected with 50 ng/well ECE and incubated with 1.25 nmol/L bigET-1 gave 40% of the maximal stimulation (Figs 3B⇑ and 5⇑). In mock-transfected or ECE-1a–transfected cells, the addition of 1 nmol/L exogenous ET-1 resulted in the same 40% of the maximal stimulation (Fig 3B⇑). This means that 1.25 nmol/L bigET-1 and 1 nmol/L ET-1 gave the same percentage of maximal stimulation, indicating that 1.25 nmol/L bigET-1 was converted into 1 nmol/L ET-1. The calculated conversion yield was therefore nearly 100%. If the same reasoning is used, it could be said that 5 nmol/L and 20 nmol/L exogenous bigET-1 gave a conversion yield of at least 40% and 15%, respectively.
Further Characterization of the Conversion of Endogenous bigET-1
A constant amount of one plasmid (preproET-1 or ECE-1a) and varying amounts of the other were cotransfected for assessing the conversion of endogenous bigET-1 (the total amount of plasmid was kept constant). Again, ECE-1a was not the limiting factor of the assay (Fig 6⇓). The same concentration range of ECE-1a plasmid gave maximal conversion of exogenous 20 nmol/L bigET-1 and of the bigET-1 generated from 250 ng/well preproET-1 plasmid. Conversely, there was no significant activity until 25 ng/well preproET-1 plasmid was transfected, and 250 ng/well preproET-1 plasmid was required to obtain maximal luciferase activity.
The same amounts of plasmid do not necessarily correspond to the same levels of expression in transfection experiments because of variations in transfection efficiency. Therefore, a constant amount of pGH110-βGal plasmid was also transfected with plasmids of interest and β-Gal activity was measured. The results showed that (1) transfection efficiency was 5-fold higher using 50 ng pcDNA3 vector compared with 50 ng pcDNA3-ECE-1a vector and that (2) conversely, transfections with pece or pece–preproET-1 always gave the same transfection efficiencies. The resulting corrections on the abscissa were not incorporated in Figs 5⇑ and 6⇑, so that they are clear and the amounts of plasmid transfected remain directly readable. The only adjustments needed for the correct interpretation of Fig 6⇑ in terms of expression are (1) a 5-fold shift to the right of the results obtained with various amounts of preproET-1 plasmid alone and (2) a slight contraction of all the curves obtained with various amounts of ECE-1a plasmid so that the minimal amount of ECE-1a plasmid is shifted 5-fold to the right and the maximal amount remains unchanged. The relative positions of the curves, showing the effects of various amounts of ECE-1a plasmid on the conversion of exogenous and endogenous bigET-1s, were therefore unchanged.
These considerations on transfection efficiency were also important for the indirect calculation of the endogenous conversion yield. Without ECE-1a, 80 and 250 ng/well preproET-1 resulted in 20% and 30% stimulation, respectively (Fig 6⇑), corresponding to the stimulation between 20% (Fig 5⇑) and 30% (Fig 3A⇑, mean of three independent experiments; Fig 3B⇑) induced by 20 nmol/L bigET-1 in mock-transfected cells. Therefore, the transfection of 80 to 250 ng/well preproET-1 resulted in the production of about 20 nmol/L bigET-1. Due to differences in transfection efficiency, more than 400 (80×5) ng/well preproET-1 plasmid would be needed when cotransfected with 50 ng/well ECE-1a plasmid to obtain the same level of expression of preproET-1 as that obtained when 80 to 250 ng/well preproET-1 plasmid was transfected alone. From Fig 5⇑, it can be assumed that 400 ng/well preproET-1 plasmid cotransfected with 50 ng/well ECE-1a would result in 100% stimulation, denoting the production of more than 10 nmol/L ET-1. Therefore, 20 nmol/L endogenous bigET-1 would result in the production of at least 10 nmol/L ET-1. Thus, the calculated conversion yield for endogenous bigET-1 is at least 50%.
Finally, an unexpected effect of pGH110-βGal plasmid aided the interpretation of the assay. This plasmid acted as a “carrier” plasmid and transfections performed with it were more efficient. Preliminary experiments performed without pGH110-βGal plasmid showed maximal stimulation of luciferase activity with greater amounts of plasmid (data not shown), indicating lower transfection efficiencies. This should be considered when comparing the results shown in Figs 3A⇑ and 6⇑. The same amounts of preproET-1 were used (250 ng/well), but different percentages of stimulation were obtained in cells transfected with preproET-1 alone (0% and 30%), and in cells transfected with ECE-1a and preproET-1 (75% and 100%). These results fit perfectly with the notion that preproET-1 is always the limiting factor.
The present study describes a novel sensitive live-cell assay for studying ECE activity. ET-1 is formed from its precursor preproET-1 via the cleavage of the intermediate bigET-1 by ECE-1. However, the subcellular site at which this step occurs is not clear: It could occur intravesicularly along the secretory pathway or bigET-1 may be released and processed extracellularly. To address this point, we have developed an integrated autocrine system.
Until now, ECE activity has been evaluated in solubilized membrane fractions by high-performance liquid chromatography assays,32 immunoassays,12 fluorogenic determinations,33 receptor assays,34 and fluorescence polarization assays.35 Most of these assays require an incubation in vitro with high concentrations of substrate, and all of them need an independent second step to measure the product, ET-1.
In the present study, a recombinant CHO reporter cell line permanently expressing the human ETA receptor and a reporter gene sensitive to its activation has been developed. This endothelin biosensor uses signaling cascades downstream from the activation of the receptor as a natural amplifier, activated by ET-1 but not by bigET-1. In addition, this system allows a significant detection of 0.05 to 10 nmol/L ET-1 comparable to that of the receptor assay.34 The reporter gene luciferase is the ultimate protagonist of the signaling cascade, and it allows simple, rapid, and linear measurement. Finally, unlike all other ECE assays, this recombinant cell line can be used to measure ECE activity directly after transient transfection of the ECE expression vector, thus allowing the entire assay to be performed on the same adherent cells in the same 96-well plates.
Our first results (Fig 3A⇑) clearly demonstrated the potential of the assay. Three characteristics of this novel assay were determined: (1) the unequivocal measurement of ET-1, as assessed by luciferase activity, must be performed between 0.1 and 10 nmol/L; (2) to allow comparison between different experiments, the results must be normalized using 10 nmol/L ET-1 as maximal stimulation (defining the 100% stimulation) and mock transfection as background (defining the 0% stimulation); and (3) the induction of luciferase by 20 nmol/L bigET-1 in the absence of ECE results in a high “background” (30%) in these cells having just undergone transient transfection. However, this background is always limited and stabilizes with increasing concentrations of bigET-1. In fact, this luciferase activity observed in the absence of ECE-1a enables the evaluation of the absolute amounts of endogenous bigET-1 produced by preproET-1–transfected cells. The effects of bigET-1 are not understood but could perhaps be explained by nonspecific cleavage of bigET-1 by CHO cells in our long-run assay. Nishikori et al36 showed that various C-terminal elongated derivatives of ET-1 bind relatively strongly to the ETA receptor and exert limited biological effects.
In the present study, we also show that phosphoramidon has different inhibitory effects on the conversion of exogenous and endogenous bigET-1s by ECE-1a. Its IC50 is 30 μmol/L for exogenous big ET-1, and more than 1 mmol/L for endogenous big ET-1 (Fig 4⇑). This was observed despite the fact that the characteristics of the two conversions are extremely similar: (1) for both conversions of bigET-1, ECE-1a is not the limiting factor, as only small amounts of ECE-1a plasmid are required; (2) in both cases, preproET-1 (endogenous bigET-1) and bigET-1 (exogenous big ET-1) are limiting for this live-cell assay; (3) the same amounts of ECE-1a plasmid are required for similar conversion activities of endogenous and exogenous bigET-1s in ET-1; and (4) the conversion yields of endogenous and exogenous bigET-1s are similar.
These results lead to two conclusions. First, they clearly demonstrate that the conversion of endogenously produced substrate to ET-1 is completed mainly intracellularly (and hence is not accessible to phosphoramidon) in CHO cells transfected with preproET-1 and ECE-1a. The second implication of our results is that the same ECE-1a molecule may account for two equally relevant activities, the intracellular and the extracellular conversions of bigET-1. These activities were indeed similar for the same levels of expression of ECE-1a.
In contrast, in the study of Xu et al,15 the two conversions did not have the same characteristics. They used a cellular source for the endogenous and the exogenous bigET-1s by transiently transfecting preproET-1 into cells stably expressing ECE-1, or by coculturing this cell line with another one stably expressing preproET-1. BigET-1 is provided by cells in both cases, but at different concentrations (300 pmol/L for endogenous bigET-1 and 1500 pmol/L for exogenous bigET-1). This may be the reason why they reported an efficacy of 5% to 10% for extracellular conversion, which is significantly lower than ours, and lower than their efficacy for the conversion of endogenous bigET-1 (50% to 90%). Indeed, we do observe that the conversion yield diminishes with increasing concentrations of bigET-1 (as explained in the “Results”), and this probably corresponds to a phenomenon of saturation. Additionally, their use of a cellular source for exogenous bigET-1 may contribute to their low yield in extracellular conversion. If the rate of bigET-1 cleavage is not much higher than its rate of production, there will always remain a high proportion of uncleaved bigET-1. Conversely, our long incubation (24 hours) and our use of a well-defined amount of bigET-1 allow ECE-1a to cleave all the bigET-1 present in the medium. Our findings show that the characteristics of the conversions of endogenous and exogenous bigET-1 were similar and that the results obtained can truly be compared.
In conclusion, the extracellular and the intracellular activities of ECE-1 appear to both be highly effective in CHO cells, each toward its own substrate. Although the physiological relevance of these activities still has to be assessed, they may explain the contradictory assumptions made about the location of ECE-1 in endothelial cells. Takahashi et al21 showed by confocal immunofluorescence microscopy that most of the ECE in endothelial cells was clustered along the plasma membrane, whereas Harrison et al22 have demonstrated by several approaches the presence of immunoreactive ET-1 in vesicular fractions within bovine endothelial cells. The location of high ECE-1 activity at two subcellular sites would have important implications for its physiological role and for the design of ECE inhibitors. First, the intracellular processing of endogenous bigET-1 in endothelial cells would imply that ECE inhibitors must be able to cross the membrane. Second, an extracellularly active ECE-1 raises the question of its physiological role if the majority of the ECE activity toward the endogenous substrate is located intracellularly in a membrane-bound compartment. This could imply either that bigETs may act as paracrine or endocrine factors activated by luminal ECE-1 (when there is an overflow of bigET-1) or that ECE-1 cleaves other circulating substrates.
Our recombinant cell line provides a simple assay for measuring ET-1 concentrations in various biological fluids under several physiological conditions, and can also be used for screening and describing ETA antagonists and agonists. Indeed, reporter cell lines are new tools for studying G protein–coupled receptors and particularly for screening agonists and antagonists.37 This study shows that a reporter cell line also allows the whole endothelin system to be studied, from the biosynthesis of the ligand to the activation of the receptor. This report is thus the first example of the use of an integrated live-cell assay to measure the activity of a processing enzyme such as ECE. It is also suitable for studying receptor activation and ET-1 biosynthesis by mutagenesis of the receptor, ECE-1, and preproET-1. Finally, a completely integrated cell line in which all protagonists, preproET-1 (under inducible form), ECE, hETA receptor, and reporter gene, are stably expressed, could be useful for screening ECE-1 inhibitors capable of entering cells.
Selected Abbreviations and Acronyms
|CHO||=||Chinese hamster ovary|
|DMEM/F-12||=||Dulbecco’s modified Eagle’s medium/Ham’s F-12|
|ET-1/ET-2/ET-3||=||endothelin-1, -2, and -3|
|hETA||=||human ETA receptor|
|TRE||=||12-o-tetradecanoyl-phorbol-13-acetate– responsive element|
This work was supported in part by the Ministère de la Recherche et de l’Enseignement (ACC-SV9). We thank Drs Jeanine Niérat and Marzia Harnois for the generous supply of oligonucleotides. We thank Dr Tracy Williams for helpful discussions. The English text was checked by Dr Owen Parkes.
- Received January 9, 1997.
- Revision received February 26, 1997.
- Accepted February 26, 1997.
Inoue A, Yanagisawa M, Kimura S, Kasuya Y, Miyauchi T, Goto K, Masaki T. The human endothelin family: three structurally and pharmacologically distinct isopeptides predicted by three separate genes. Proc Natl Acad Sci U S A. 1989;86:2863-2867.
Opgenorth T, Wu-Wong J, Shiosaki K. Endothelin-converting enzymes. FASEB J. 1992;6:2653-2659.
Kimura S, Kaysuya Y, Sawamura T, Shinmi O, Sugita Y, Yanagisawa M, Goto K, Masaki T. Conversion of big endothelin-1 to 21-residue endothelin-1 is essential for expression of full vasoconstrictor activity: structure-activity relationships of big endothelin-1. J Cardiovasc Pharmacol. 1989;13:S5-S7.
McMahon E, Palomo M, Moore W, McDonald J, Stern M. Phosphoramidon blocks the pressor activity of porcine big endothelin-1-(1-39) in vivo and conversion of big endothelin-1-(1-39- to endothelin-1-(1-21) in vitro. Proc Natl Acad Sci U S A. 1991;88:703-707.
Ahn K, Beningo K, Olds G, Hupe D. The endothelin-converting enzyme from human umbilical vein is a membrane-bound metalloprotease similar to that from bovine endothelial cells. Proc Natl Acad Sci U S A. 1992;89:8606-8610.
Shimada K, Takahashi M, Tanzawa K. Cloning and functional expression of endothelin-converting enzyme from rat endothelial cells. J Biol Chem. 1994;269:18275-18278.
Valdenaire O, Rohrbacher E, Mattei MG. Organization of the gene encoding the human endothelin-converting enzyme (ECE-1). J Biol Chem. 1995;270:29794-29798.
Takahashi M, Fukuda K, Shimada K, Barnes K, Turner AJ, Ikeda M, Koike H, Yamamoto Y, Tanzawa K. Localization of rat endothelin-converting enzyme to vascular endothelial cells and some secretory cells. Biochem J. 1995;311:657-665.
Harrison V, Barnes K, Turner A, Wood E, Corder R, Vane J. Identification of endothelin-1 and big endothelin-1 in secretory vesicles isolated from bovine aortic endothelial cells. Proc Natl Acad Sci U S A. 1995;92:6344-6348.
Itoh Y, Yanagisawa M, Ohkubo S, Kimura C, Kosaka T, Inoue A, Ishida N, Mitsui Y, Onda H, Fujino M, Masaki M. Cloning and sequence analysis of cDNA encoding the precursor of a human endothelium-derived vasoconstrictor peptide, endothelin: identity of human and porcine endothelin. FEBS Lett. 1988;231:440-444.
Ohlstein EH, Nambi P, Douglas SA, Edwards RM, Gellai M, Lago A, Leber JD, Cousins RD, Gao A, Frazee JS. SB209670, a rationally designed potent nonpetide endothelin receptor antagonist. Proc Natl Acad Sci U S A. 1994;91:8052-8056.
Graham F, VanDerEb A. A new technique for the assay of infectivity of human adenovirus DNA. Virology. 1975;52:456-476.
Aramori I, Nakanishi S. Coupling of two endothelin receptor subtypes to different signal transduction in transfected chinese hamster ovary cells. J Biol Chem. 1992;267:12468-12474.
Sawamura T, Kasuya Y, Matsushita Y, Suzuki N, Shinmi O, Kishi N, Sugita Y, Yanagisawa M, Goto K, Masaki T, Kimura S. Phosphoramidon inhibits the intracellular conversion of big endothelin-1 to endothelin-1 in cultured endothelial cells. Biochem Biophys Res Commun. 1991;176:860-865.
Takahashi M, Matsushita Y, Iijima Y, Tanzawa K. Purification and characterization of endothelin-converting enzyme from rat lung. J Biol Chem. 1993;268:21394-21398.