- Email: [email protected]
A permanent human cell line (EA.hy926) preserves the characteristics of endothelin converting enzyme from primary human umbilical vein endothelial cells
Life Sciences, Vol. 56, No. 26 pp. 2331-2341,
Pergamon
199.5
Copyright 0 1995 ESevier Science Ltd Printed in the USA.
All rights resexvcd
0024-3205/95
$950
t .oo
0024-3205(95)00227-8
A PERMANENT HUMAN CELL LINE (EA.hy926) PRESERVES THE CHARACTERISTICS OF ENDOTHELIN CONVERTING ENZYME FROM PRIMARY HUMAN UMBILICAL VEIN ENDOTHELIAL CELLS
Kyunghye
Ahn*, Sharon Pan, Karen Beningo, and Donald Hupe
Department of Biochemistry, Parke-Davis Pharmaceutical Research, Division of Warner-Lambert Company, Ann Arbor, Michigan 48 105 (Received in final form March 25, 1995)
Summary
Purification of endothelin converting enzyme (ECE) from endothelial cells has been hindered by the difficulty in obtaining primary endothelial cells in large quantity. We therefore tested transformed human umbilical vein endothelial cells (EA.hy 926) for ECE activity. Our data clearly demonstrate that this transformed cell line preserves the ECE properties of the primary cell line. These include: (i) one sharp activity optimum at neutral pH; (ii) characteristics typical of a metalloprotease; (iii) IC50 value for phosphoramidon of 1.8 l_tM (2.7 pM for HUVEC); (iv) no inhibition by captopril and thiorphan, inhibitors of angiotensin converting enzyme and neutral endopeptidase 24.11. The enzyme showed a substrate specificity for big ET-1:big ET-2:big ET-3 in a ratio of 40:2.5: 1. This report presents evidence that a permanent human endothelial cell line, EA.hy926, preserves the ECE activity of HUVEC and is useful for the study of ECE and its regulation of ET-l production. Key Wurak: endothelin converting enzyme, umbilical vein, human cell
line
Endothelin-1 (ET-l) is a peptide with potent vasoconstrictor activity first isolated from the medium of cultured porcine aortic endothelial cells (1). Based upon sequence analysis of cDNA for ET-l, Yanagisawa et al. proposed that ET- 1 is generated by an unusual processing reaction involving a cleavage between Trp21 and Va122 of big endothelin-1 (big ET-l) by a putative endothelin converting enzyme (ECE) (1). Since the vasoconstrictor activity of big ET-l is much lower than that of ET-1 (2,3), the conversion from big ET-l to ET-l appears to be essential for physiological activity. Indeed, it has been demonstrated that intravenous injection of big ET-l in the rabbit actually resulted in increased circulating levels of ET-l (4). The many biological functions for ET-l have been reviewed (56). Because phosphoramidon has been shown to inhibit the pressor and airway contractile effects of big ET-l in vivo (7, 8) and to suppress the secretion of ET-l from cultured endothelial cells (9, *Corresponding author: Kyunghye Ahn, Department of Biochemistry, Parke-Davis Pharmaceutical Research, Division of Warner-Lambert Company, Ann Arbor, Michigan
48105
2332
Endothelin Converting Enzyme
Vol. 56, No. 26,199s
lo), a phosphoramidon-inhibitable ECE has been postulated. In contrast, inhibitors of cathepsin E and angiotensin converting enzyme did not inhibit the big-ET induced pressor response (11, 12). Consistent with these observations in vivo, several research groups have reported ECE activities from various sources that were inhibitable by phosphoramidon but not by inhibitors of angiotensin converting enzyme (13-15). Recently, reports on the cloning of rat and bovine ECE have been published (16-18). We have shown that ECE from human umbilical vein endothelial cells (HUVEC) is a membranebound and phosphoramidon-inhibitable metalloprotease (15). Human ECE was of particular interest because of its implication as an important therapeutic target for cardiovascular diseases. Further purification, however, has been hindered because of difficulty in obtaining primary endothelial cells in large quantity. This study therefore was initiated to identify a transformed human cell line which would preserve all the characteristics of ECE without lowering the expression level of ECE compared to its primary cell source. In this study, transformed HUVEC (EA.hy 926) were tested for ECE activity. Our data demonstrated that this cell line preserves all the characteristics of ECE from primary cells with a specific activity similar to that of HUVEC.
Materials and Methods
Materials: Human big ET-l (l-38), big ET-2 (l-37), big ET-3 (1-41 amide), ET-l (l-21), ET-2 (l-21), and ET-3 (1-21) were purchased from Peptide International (Louisville, KY). [125I]ET-1 was from Amersham. Human umbilical vein endothelial cells and media for their growth were from Clonetics (San Diego, CA). Polyclonal antibody against ET-l was from Biodesign International (Kennebunkport, ME). Phosphoramidon, pepstatin A, and leupeptin were from Boehringer Mannheim. E-64, l,lO-phenanthroline, thiorphan, bestatin, and soy bean trypsin inhibitor were obtained from Sigma. Triton X-100 (hydrogenated) and Tween 20 were from Calbiochem. Ricinus communis agglutinin (RCA)-1 resin was from EY Laboratories, Inc (San Mateo, CA). Monoclonal antibody against neutral endopeptidase 24.11 (NEP 24.11) [also known as common acute lymphoblastic leukemia antigen (CALLA)], J5, was from Coulter (Hialeah, FL). Rabbit kidney NEP 24.11 was obtained from Dr. P. Crine (Department of Biochemistry, University of Montreal, Quebec). Protein A-sepharose was obtained from Pharmacia LKB Biotechnology Inc. Buffers: Buffer A is 60 mM KPi, pH 7.4/10 mM EDTA/8 mM NaN3/0.1% bovine serum albumin/O.l% Tween 20; buffer B is 60 mM KPi, pH 7.4/10 mM EDTA/8 mM NaN3/O.25% gelatin (wt/vol); buffer C is 20 mM Tris-HCl, pH 7.5/0.02% NaN3; buffer D is 10 mM Tris-HCl, pH 7.5/0.25 M sucrose/20 mM KCl; buffer E is 50 mM Tris-HCl, pH 7.2150 mM NaCl/0.02% Protease inhibitors cocktail is 1 mM NaN3/0.2% Triton X-100 (hydrogenated); phenylmethylsulfonyl fluoride/O.05 mM pepstatin A/O. 1 mM leupeptin. Measurement of Endothelin-1 Converting Enzyme Activity: The assay measured the production of ET- 1 essentially as described with minor modifications (15). The reaction mixture (50 l.tl) contained 10 l.tM big ET-l, 100 mM Hepes-KOH (pH 7.0), 0.25% Triton X-100, 0.01% NaN3,0.2 mM phenylmethlysulfonyl fluoride, 0.02 mM pepstatin A, 0.1 mM leupeptin, and the enzyme as indicated. After incubation for the indicated period stopped by the addition of EDTA to give a final concentration diluted with buffer A and the generated ET-l was measured RIA mixture (250 ~1) contained the antibody against ET-l, (15000 cpm) in buffer A. The order of the addition was ET-
of time at 37 OC, the reaction was of 10 r&I. This final mixture was by radioimmunoassay (RIA). The an ET-l sample, and [ l251] ET-l 1 sample, followed by antibody, and
Vol. 56, No. 26, 1995
Endothelin Converting Enzyme
2333
then [125I]ET-1.
After incubation at 4 oC for 16 hr, unbound ET-l was co-precipitated by the addition of charcoal (2.4%, wt/vol)/dextran (0.24%, wt/vol) suspension (125 p.1) in buffer B. The amount of ir-ET-l was measured by counting the supernatant and determined from the standard curve. The cross reactivity to big ET-l was less than 0.01% and the detection limit was 1 fmol. One unit of enzyme is defined as the amount generating 1 pmol of immunoreactive (ir) ET-l per min. Cell Culture: BAEC and HUVEC were grown as described (15). The typical enzyme preparation was performed after obtaining confluent cells at 5 to 8 passages from 50-100 roller bottles. The membrane fractions were prepared essentially the same as before (15). A permanent human cell line (EA.hy926), derived by fusing human umbilical vein endothelial cells with the permanent human cell line A549 (derived from a human lung carcinoma), was obtained from Dr. C.-J. S. Edge11 (Department of Pathology, University of North Carolina at Chapel Hill, NC) and was grown as described (19). Immunoprecipitation: The mixture (270 pl) contained protein A-sepharose (70 l,tl) which was equilibrated with 50 mM Hepes-KOH (pH 7.4), and J5 (50 pg), a monoclonal antibody against neutral endopeptidase 24.11 (NEP 24.11). After incubation for 20 h at 4 oC, the mixture was centrifuged for 5 min at 10,000 x g. The pellet was then washed five times with 50 mM HepesKOH (pH 7.4) containing 3% BSA and incubated for 1 day at 4 oC either with RCA-I purified enzyme (23.7 pg) from EA.hy926 or rabbit kidney NEP 24.11 (1.7 pg) in buffer C containing 0.5% Triton X-100 (hydrogenated) and protease inhibitors cocktail (total volume 500 l.~l). After centrifugation as above, the supernatant (36.5 pl) was assayed for the enzyme activity as described above. Partial Purification of ECE from EA.hy926: All operations were carried out at O-4 oC unless otherwise noted. The cells in each roller bottle were washed with phosphate-buffered saline and gently scraped. These cells were washed further with phosphate-buffered saline followed by buffer D and frozen immediately in liquid nitrogen. The cells from 150 roller bottles were suspended in 100 ml of buffer D containing protease inhibitors cocktail and were homogenized via nitrogen cavitation (600 psi, 10 min.) and centrifuged at 5,000 x g for 20 min. This process was repeated with the pellet resuspended in 100 ml of buffer D containing protease inhibitors cocktail. The combined supernatant was then centrifuged at 20,000 x g for 35 min. The resulting supematant was further centrifuged at 100,000 x g for 1 h. The pellet was washed with 120 ml of buffer C, resuspended in 40 ml of buffer C containing 0.5% Triton X-100 (hydrogenated) and protease inhibitors cocktail, and was stirred gently for 1 h (membrane fraction). The clear supematant was obtained by centrifugation at 100,000 x g for 1 h (detergent extract). For RCA-I, the detergent extract was applied at a flow rate of 0.15 ml/mm onto a 4-ml RCA-I column (0.5 x 20 cm) equilibrated with buffer E including protease inhibitors cocktail. The column was washed with the equilibration buffer until A280 nm of the eluate was less than 0.03, and the activity was eluted with buffer E containing 0.5 M galactose at a flow rate of 0.15 ml/min. Fractions of 4 ml were collected and peak fractions were pooled (RCA-I fraction, 20 ml).
Results EA.hy926 Possesses Two Endothelin-1 Converting Activities. When endothelin-1 converting activity from the membrane fraction was tested in the presence of phosphoramidon across a wide range of concentrations (0.1 nM - 60 /.tM), a biphasic curve was produced, consistent with the inhibition of two distinct enzyme species with two different affinities for phosphoramidon (Fig. 1). The resulting data set could be fit to two separate IC50 curves,
Endothelin
2334
Converting Enzyme
Vol. 56, No. 26, 1995
giving IC50 values of 2.5 f 0.45 nM (activity I) and 1.5 k 0.48 pM (activity II). Detergent extraction followed by RCA-I chromatography gave 38.5- and 32.4-fold purification for total endothelin-1 converting activity and activity II (Table I). However, as indicated by the biphasic curve in the presence of phosphoramidon, this procedure did not separate the two activities (data not shown).
2.5
-
1.5
-
1
-
= E e 7 iTI C
0.5
-
lo-‘*
IO.6
10-E
lo-‘0
[Phosphoramidon]
1 o-4
(M)
Fig. 1 Presence of two endothelin-1 converting activities in EA.hy926. The membrane fraction (5.9 pg) from EA.hy926 was incubated at 37 oC for 1 hr with the indicated phosphoramidon concentrations. The amount of ET- 1 generated was analyzed by RIA.
Table I Partial Purification
of ECE from EA.hy926a Total Activity
Fraction
Total Volume
ml
Membrane Detergent extract RCA-I
Protein Concentration
Absorbance(280
I+II
nm)
IIb
units
Specific Activity 1+11
nb
unitdmVA2gO
nm
40.0 39.4
0.902c 0.387c
1324 1159
382 301
3.7 7.6
1.1 2.0
20.0
0.386
1101
275
142.6
35.6
aECE was partially purified from EA.hy926 grown in 150 roller bottles to con fluency. bActivity was measured in the presence of 100 ~_LM thiorphan, which completely inhibited activity I (see “Results”). cAbsorbance shown was measured after lo-fold dilution.
Vol. 56, No. X,1995
EndothelinConvertingEnzyme
lo-8 [Thlorphan]
2335
1 o-7 (M)
Fig. 2 Inhibition of activity I by thiorphan. The detergent-extracted fraction (2.2 pg) from EA.hy926 was incubated at 37 OC for 1 hr with the indicated thiorphan concentrations. The amount of ET- 1 generated was analyzed by RIA. Activity I, Which Has an Affinity For Phosphoramidon In the nM Range, Is Due To Neutral Endopeptidase 24.11 (NEP 24.11). It has been suggested that NEP 24.11 cleaves big ET- 1 at several sites including at the Trp-2 1-Val22 bond (20). Vijayaraghavan et al. also reported that NEP 24.11 can efficiently hydrolyze ET-l with a kcat value of 13 1.4 mitt-1 (2 1). Recently Murphy et al. reported the relative half-lives for hydrolysis of ET-I, big ET-I, and C-terminal fragment (22-38) by NEP 24.11 (22). When the inhibition of endothelin-1 converting activity from the detergent extract was tested in the presence of thiorphan, a NEP 24.11 inhibitor with a Ki value of 2.2 nM (23), an IC50 value of 42.4 + 10.9 nM was obtained (Fig. 2). Also, immunoprecipitation of the RCA-I fraction with J5, a monoclonal antibody against human NEP 24.11, showed 80 % inhibition of endothelin- 1 converting activity. In contrast, rabbit kidney NEP 24.11 did not coprecipitate with J5 (Table II). These data, including IC50 values of 2.5 and 42.4 nM for phosphoramidon and thiorphan, respectively, and coprecipitation of the activity with a monoclonal antibody against human NEP 24.11, indicate that the activity I discussed above is due to NEP 24.11. Activity II, Which Has an Affinity For Phosphoramidon In the p.M Range, Is Similar To ECE Activities From HUVEC And BAEC. In order to test the inhibition of activity II from EA.hy926 by thiorphan, the enzyme reaction was performed in the presence of 100 nM phosphoramidon. Under these conditions, activity I should be completely inhibited (Fig. 1). Thiorphan, at concentrations up to 100 pM, did not inhibit activity II. When phosphoramidon inhibition of activity II was measured in the presence of 100 pM thiorphan in order to completely inhibit activity I, an IC50 value of 1.8 + 0.21 p.M was obtained (Fig 3). This value is comparable to those from the primary cell lines, HUVEC (2.7 pM) and BAEC (1.8 pM) (15). Other characteristics of activity II which are similar to ECE activities from HUVEC and BAEC include the following: Activity II was abolished by phosphoramidon, EDTA, and l,lO-phenanthroline but was not inhibited by pepstatin A, phenylmethlysulfonyl fluoride, soybean trypsin inhibitor, leupeptin, or E-64 - consistent with
Endothelin
2336
Converting
Vol. 56, No. 26,1995
Enzyme
Inhibitors of cathepsin E (SQ 32056) and angiotensin characteristics of a metalloprotease. converting enzyme (captopril and enalaprilat) did not inhibit activity II. Bestatin, an inhibitor for various aminopeptidases, did not affect the activity (Table III). The pH optimum was also narrow with a maximum at neutral pH (data not shown). Additionally, the specific ECE activity of 5.4 units/ml/mg membrane fraction from EA.hy926 (Table I) was comparable to that from primary HUVEC (3.9 units/ml/mg membrane fraction).
Table II Precipitation
of NEP 24.11 Activity by a Monoclonal Endothelin-1
Enzyme
Antibody, J5a
Converting Activity
(% of Control) RCA-I fraction from EA.hy926
19.9 f 2.7
Rabbit kidney NEP 24.11
90.5 + 8.5
aIn each experiment, the indicated enzyme was incubated for 1 day with J5 bound to protein A-sepharose as described in “Materials and Methods.” After centrifugation, the enzyme activity remaining in the supernatant was measured. All experiments were carried out in duplicates. The average activity is expressed as a percentage of the control (enzyme incubated with protein A-sepharose alone).
l -x.
.
-
Fig. 3 Inhibition of ECE activity (activity II) by phosphoramidon. The RCA-I fraction (1.9 pg) from EA.hy926 was incubated in the presence of 100 PM thiorphan at 37 oC for 1 hr with the indicated phosphoramidon concentrations. The amount of ET-l generated was analyzed by RIA.
Vol. 56, No. 26, 1995
Endothelin
Converting
Enzyme
2337
Table III Inhibition
Profile of ECE Activity from EA.hy926 Concentration
Inhibitors
ECE Activity a (% of Control)
Phosphoramidon
50 pM
EDTA
50 pM
1, 10-Phenanthroline
100 pM
73.6
500 pM
4.0
6.3 6.4
100 pM
112.6
500 pM
8.6
Captopril
100 pM
97.9
Enalaprilat
100 pM
121.0
SQ 32056
100 pM
98.7
200 pM
130.7
Soybean trypsin inhibitor
0.13 mg/ml
150.4
Pepstatin A
20 pM
136.2
Leupeptin
200 pM
135.7
E-64
500 pM
140.0
Bestatin
150 pM
95.8
EGTA (2 mM)
Phenylmethlysulfonyl
fluoride
aThe RCA-I fraction (2.6 pg) of EA.hy926 was incubated in the presence of 100 pM thiorphan at 37 oC for 1 hr with or without the indicated protease inhibitors. The amount of ET-l generated was analyzed by RIA. The amount of ir-ET-l in the control was 7.4 pmol. The final concentration of organic solvent (dimethyl sulfoxide, isopropyl alcohol, ethanol, or methanol) was 1%.
For comparison, endothelin-1 converting activity of the membrane fraction from HUVEC was tested in the presence of thiorphan (0.01-100 PM). These conditions produced approximately IO20 % inhibition compared to the control (without thiorphan), suggesting that at most lo-20 % of the total endothelin-1 converting activity is due to NEP 24.1 I in HUVEC (data not shown). The combined results indicate that the level of NEP 24.11 activity in HUVEC is much lower than that in the hybrid cells, EA.hy926. The expression of NEP 24.11 in HUVEC has been reported (24). Z& Inhibits ECE Activity. When additional ZnC12 was added to the assay, the ECE activity
was inhibited
in a dose-
dependent manner with an IC50 value of 3.2 * 0.2 pM (data not shown). Similar Zn2+ sensitivity was also observed with the ECE activity from HUVEC (IC50: 10.8 + 0.8 pM) (data not shown). At 10 pM, neither M&12, MgC12, CaC12, NiC12, nor CoC12 had any effect.
Endothelin Converting Enzyme
2338
Vol. 56, No. 26, 1995
Substrate Specificity. ECE showed a high specificity for big ET- 1. When ECE was incubated with big ET-l, big ET-2, or big ET-3, it showed a specificity ratio of 40:2.5:1 for big ET-1:big ET-2:big ET-3 (Fig. 4). Similar specificity was also observed with the ECE activity from HUVEC (data not shown). Sawamura et al. reported similar substrate specificity for their thiorphan-insensitive ECE activity from porcine lung membrane (20). 1.2
1
2 Time
3
(hr)
Fig. 4 Substrate specificity of ECE. The RCA-l fraction (2.3 l.tg) from EA.hy926 was incubated with 1 pM big ET-l (a), big ET-2 (0), or big ET-3 (A) in the presence of 100 pM thiorphan for the indicated period of time. The product produced was quantified by RIA from standard curves generated using known amounts of ET-l , ET-2, or ET-3. Discussion
Several groups have reported phosphoramidon-inhibitable ECE activities from various sources (13-15,20,25-29). It is noteworthy that the majority of these ECE activities gave IC50 values in the low pM range for phosphoramidon regardless of the source (14-15,20,27-29). One can use this level of phosphoramidon sensitivity (ie: PM range) as a tool to differentiate genuine ECE activity from other converting activities, assuming that this is a defining property of ECE. Other e.g., angiotensin converting enzyme, NEP 24.11, phosphoramidon-sensitive enzymes, carboxypeptidase A, P. aeruginosa elastase, and thermolysin, are known to have much tighter binding for phosphoramidon (in the nM range of IC50) (30). We therefore used this characteristic of phosphoramidon sensitivity in the pM range to search for ECE activity from the transformed human endothelial cell line, EA.hy926, and showed that there are two endothelin-1 converting activities based on the biphasic curve. It was demonstrated that the endothelin-1 converting activity with an IC50 value of 2.5 nM for phosphoramidon is due to the presence of NEP 24.11. As shown in Fig. 2, when thiorphan inhibition of endothelin-1 converting activity from the detergent extract of EA.hy926 was tested, a maximum inhibition level of approximately 75-80% was reached at about 1 l.tM and remained at
Vol. 56, No. X,1995
Endothelin Converting Enzyme
2339
this level even up to 100 PM (Fig. 2). This same level of inhibition was seen for the first curve in the biphasic phosphoramidon titration (Fig. 1) and when RCA-I purified ECE from EA.hy926 was immunoprecipitated with a monoclonal antibody against human NEP 24.11 (Table II). These results indicate that ECE activity is approximately 20-25% of the total endothelin-1 converting activity from EA.hy926. While this paper was in preparation, the use of this cell line for studying ECE was also described by Waxman et al. (3 1). They reported that the human lung carcinoma A549 cells contain less than 10 % of the ECE activity present in EA.hy926 cells. Since the ECE activity from the primary HUVEC is 80-90 % of total endothelin-1 converting activity and is comparable to that from the hybrid cells, the overexpression of NEP 24.11 in EA.hy926 cells was probably a consequence of the hybridization procedure. Others have previously suggested that NEP 24.11 is expressed in EA.hy926 (32, 33). The current study indicates that the ECE activity of HUVEC has been preserved in EA.hy926. By completely inhibiting the NEP 24.11 activity in EA.hy926 using thiorphan in a manner that does not affect the ECE activity, the ECE activity has been successfully characterized based on pH dependence, protease inhibitor effect, and substrate specificity. We also demonstrated that the ECE activity is inhibited by additional Zn2+ with 1C50 values of 31 1 l.tM. Waxman et al. reported the effect of ZnC12 on EDTA-treated and completely inactivated enzyme. The maximal ECE activity was obtained at 50 l.tM ZnC12 but a higher concentration (100 PM) was inhibitory (31). This difference in the inhibitory ranges of ZnC12 is likely due to the different residual ZnC12 concentrations in partially purified enzymes. High concentrations of Zn2+ (in the mM range) are known to often inhibit metalloproteases due to the formation of zinc monohydroxide that bridges the catalytic zinc ion to a side chain in the active site of the enzyme (34). Besides metalloproteases there are some cases where Zn 2+ inhibition has been observed. The cysteine protease, calpain, and the acid protease, HIV protease, are both also known to be inhibited by Zn2+ in the PM range (Wang, K., personal communication) (35). Further study might be useful to elucidate the mechanism of this Zn2+ inhibition of ECE activity. It has been shown here that the specific ECE activity obtained from EA.hy926 is in the same range as that from primary HUVEC (See “Results”). Therefore one can utilize this cell line for ECE purification without sacrificing the specific activity. This study demonstrates that a permanent human endothelial cell line, EA.hy926, useful for the study of ECE and its regulation of ET-l production.
could be very
Acknowledgments
We thank
Kenneth
Ray III and Gregory
Olds for assistance
in cell
culture
and enzyme
assays.
References
1.
2. 3.
M. YANAGISAWA, H. KURIHARA, S. KIMURA, Y. TOMOBE, M. KOBAYASHI, Y. MITSUI, Y. YAZAKI, K. GOT0 and T. MASAKI, Nature 332 411-415 (1988). T. KASHIWABARA, Y. INAGAKI, H. OHTA, A. IWAMATSU, M. NOMIZU, A. MORITA and K. NISHIKORI, FEBS Lett. 247 73-76 (1989). S. KIMURA, Y. KASUYA, T. SAWAMURA, 0. SHINMI, Y. SUGITA, M. YANAGISAWA, K. GOT0 and T. MASAKI, J. Cardiovasc. Pharmacol. 13 ISUDDI. 5) S5-S8 (1989).
234a
4. 5. 6. 7. 8. 9. 10.
11.
12. 13. 14. 15. 16. 17. 18.
19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
Endothelin Converting Enzyme
Vol. 56, No. 26, 1995
P. D’ORLEANS-JUST, P. S. LIDBURY, T. D. WARNER and J. R. VANE, Biochem. Pharmacol. 39 R21-R22 (1990). G. M. RUBANYI and L. H. BOTELHO, FASEB J. 5 2713-2720 (1991). A. M. DOHERTY, J. Med. Chem. 35 1493-1508 (1992). Y. MATSUMURA, K. HISAKI, M. TAKAOKA and S. MORIMOTO, Eur. J. Pharmacol. 185 103-106 (1990). T. FUKURODA, K. NOGUCHI, S. TSUCHIDA, M. NISHIKIBE, F. IKEMOTO, K. OKADA and M. YANO, Biochem. Biophys. Res. Commun. 172 390-395 (1990). R. IKEGAWA, Y. MATSUMURA, Y. TSUKAHARA, M. TAKAOKA and S. MORIMOTO, Biochem. Biophys. Res. Commun. 171669-675 (1990). T. SAWAMURA, Y. KASUYA, Y. MATSUSHITA, N. SUZUKI, 0. SHINMI, N. KISHI, Y. SUGITA, M. YANAGISAWA, K. GOTO, T. MASAKI and S. KIMURA, Biochem. Biophys. Res. Commun. 174 779-84 (1991). J. E. BIRD, T. L. WALDRON, D. K. LITTLE, M. M. ASAAD, C. R. DORSO, G. DIDONATO and J. A. NORMAN, Biochem. Biophys. Res. Commun. 182 224-231 (1992). D. M. POLLOCK and T. J. OPGENORTH, Am. J. Physiol. &&I R257-R263 (1991). K. OKADA, Y. MIYAZAKI, J. TAKADA, K. MATSUYAMA, T. YAMAKI and M. YANO, Biochem. Biophys. Res. Commun. 171 1192-l 198 (1990). Y. HIOKI, K. OKADA, H. ITO, K. MATSUYAMA and M. YANO, Biochem. Biophys. Res. Commun. 174 446-45 1 (199 1). K. AHN, K. BENINGO, G. OLDS and D. HUPE, Proc. Natl. Acad. Sci. USA. Sq 86068610 (1992). K. SHIMADA, M. TAKAHASHI and K. TANZAWA, J. Biol. Chem. 269 18275-18278 (1994). D. XU, N. EMOTO, A. GIAID, C. SLAUGHTER, S. KAW, D. DEWIT and M. YANAGISAWA, Cell a,473-485 (1994). T. IKURA, T. SAWAMURA, T. SHIRAKI, H. HOSOKAWA, T. KIDO, H. HOSHIKAWA, K. SHIMADA, K. TANZAWA, S. KOBAYASHI, S. MIWA and T. MASAKI, Biochem. Biophys. Res. Commun. 203 1417-1422 (1994). C.-J. S. EDGELL, C. C. MCDONALD and J. B. GRAHAM, Proc. Natl. Acad. Sci. USA. 85) 3734-3737 (1983). T. SAWAMURA, 0. SHINMI, N. KISHI, Y. SUGITA, M. YANAGISAWA, K. GOTO, T. MASAKI and S. KIMURA, Biochim. Biophys. Acta 1161295-302 (1993). J. VIJAYARAGHAVAN, A. G. SCICLI, 0. A. CARRETERO, C. SLAUGHTER, C. MOOMAW and L. HERSH, J. Biol. Chem. 265 14150-14155 (1990). L. J. MURPHY, R. CORDER, A. I. MALLET and A. J. TURNER, Br. J. Pharmacol. 113, 137-142 (1994). A. BEAUMONT, B. BARBE, H. L. MOUAL, G. BOILEAU, P. CRINE, M.-C. FOURNIE-ZALUSKI and B. P. ROQUES, J. Biol. Chem. 267 2138-2141 (1992). C. LLORENS-CORTES, H. HUANG, P. VICART, J.-M. GASC, D. PAULIN and P. CORVOL, J. Biol. Chem. 267 14012-14018 (1992). K. OHNAKA, R. TAKAYANAGI, T. YAMAUCHI, H. OKAZAKI, M. OHASHI, F. UMEDA and H. NAWATA, Biochem. Biophys. Res. Commun. 168 1128-l 136 (1990). Y. MATSUMURA, Y. TSUKAHARA, K. KUNINOBU, M. TAKAOKA and S. MORIMOTO, FEBS Lett. 305 86-90 (1992). K. OHNAKA, M. NISHIKAWA, R. TAKAYANAGI, M. HAJI and H. NAWATA, Biochem. Biophys. Res. Commun. 185 611-616 (1992). M. TAKAHASHI, Y. MATSUSHITA, Y. IIJIMA and K. TANZAWA, J. Biol. Chem. 268 21394-21398 (1993). K. OHNAKA, R. TAKAYANAGI, M. NISHIKAWA, M. HAJI and H. NAWATA, J. Biol. Chem. 268 26759-26766 (1993). J. C. POWERS and J. W. HARPER, Protease Inhibitors, A. J. Barrett and G. Salvesen (Eds), 219-289, Elsevier, Amsterdam (1986).
Vol. 56, No. 26, 1995
31. 32. 33. 34. 35.
Endothelin Converting Enzyme
L. WAXMAN, K. P. DOSH, S. L. GAUL, S. WANG, R. A. BEDNAR and A. M. STERN, Arch. Biochem. Biophys. 308 240-254 (1994). 0. SAIJONMAA, T. NYMAN, U. HOHENTHAL and F. FYHRQUIST, Biochem. Biophys. Res. Commun. 181529-536 (1991). R. CORDER, V. J. HARRISON, N. KHAN, E. ANGGARD, and J. R. VANE, J. Cardiovasc. Pharmacol. 22 (SUDD~.8) S73-S76 (1993). K. S. LARSEN and D. S. AULD, FASEB J. 2 4754 (1988). Z.-Y. ZHANG, I. M. REARDON, J. 0. HUI, K. L. O’CONNELL, R. POORMAN, TOMASSELLI and R. HEINRIKSON, Biochemistry 30 8717-8721 (1991).
2341
A.
Pergamon
199.5
Copyright 0 1995 ESevier Science Ltd Printed in the USA.
All rights resexvcd
0024-3205/95
$950
t .oo
0024-3205(95)00227-8
A PERMANENT HUMAN CELL LINE (EA.hy926) PRESERVES THE CHARACTERISTICS OF ENDOTHELIN CONVERTING ENZYME FROM PRIMARY HUMAN UMBILICAL VEIN ENDOTHELIAL CELLS
Kyunghye
Ahn*, Sharon Pan, Karen Beningo, and Donald Hupe
Department of Biochemistry, Parke-Davis Pharmaceutical Research, Division of Warner-Lambert Company, Ann Arbor, Michigan 48 105 (Received in final form March 25, 1995)
Summary
Purification of endothelin converting enzyme (ECE) from endothelial cells has been hindered by the difficulty in obtaining primary endothelial cells in large quantity. We therefore tested transformed human umbilical vein endothelial cells (EA.hy 926) for ECE activity. Our data clearly demonstrate that this transformed cell line preserves the ECE properties of the primary cell line. These include: (i) one sharp activity optimum at neutral pH; (ii) characteristics typical of a metalloprotease; (iii) IC50 value for phosphoramidon of 1.8 l_tM (2.7 pM for HUVEC); (iv) no inhibition by captopril and thiorphan, inhibitors of angiotensin converting enzyme and neutral endopeptidase 24.11. The enzyme showed a substrate specificity for big ET-1:big ET-2:big ET-3 in a ratio of 40:2.5: 1. This report presents evidence that a permanent human endothelial cell line, EA.hy926, preserves the ECE activity of HUVEC and is useful for the study of ECE and its regulation of ET-l production. Key Wurak: endothelin converting enzyme, umbilical vein, human cell
line
Endothelin-1 (ET-l) is a peptide with potent vasoconstrictor activity first isolated from the medium of cultured porcine aortic endothelial cells (1). Based upon sequence analysis of cDNA for ET-l, Yanagisawa et al. proposed that ET- 1 is generated by an unusual processing reaction involving a cleavage between Trp21 and Va122 of big endothelin-1 (big ET-l) by a putative endothelin converting enzyme (ECE) (1). Since the vasoconstrictor activity of big ET-l is much lower than that of ET-1 (2,3), the conversion from big ET-l to ET-l appears to be essential for physiological activity. Indeed, it has been demonstrated that intravenous injection of big ET-l in the rabbit actually resulted in increased circulating levels of ET-l (4). The many biological functions for ET-l have been reviewed (56). Because phosphoramidon has been shown to inhibit the pressor and airway contractile effects of big ET-l in vivo (7, 8) and to suppress the secretion of ET-l from cultured endothelial cells (9, *Corresponding author: Kyunghye Ahn, Department of Biochemistry, Parke-Davis Pharmaceutical Research, Division of Warner-Lambert Company, Ann Arbor, Michigan
48105
2332
Endothelin Converting Enzyme
Vol. 56, No. 26,199s
lo), a phosphoramidon-inhibitable ECE has been postulated. In contrast, inhibitors of cathepsin E and angiotensin converting enzyme did not inhibit the big-ET induced pressor response (11, 12). Consistent with these observations in vivo, several research groups have reported ECE activities from various sources that were inhibitable by phosphoramidon but not by inhibitors of angiotensin converting enzyme (13-15). Recently, reports on the cloning of rat and bovine ECE have been published (16-18). We have shown that ECE from human umbilical vein endothelial cells (HUVEC) is a membranebound and phosphoramidon-inhibitable metalloprotease (15). Human ECE was of particular interest because of its implication as an important therapeutic target for cardiovascular diseases. Further purification, however, has been hindered because of difficulty in obtaining primary endothelial cells in large quantity. This study therefore was initiated to identify a transformed human cell line which would preserve all the characteristics of ECE without lowering the expression level of ECE compared to its primary cell source. In this study, transformed HUVEC (EA.hy 926) were tested for ECE activity. Our data demonstrated that this cell line preserves all the characteristics of ECE from primary cells with a specific activity similar to that of HUVEC.
Materials and Methods
Materials: Human big ET-l (l-38), big ET-2 (l-37), big ET-3 (1-41 amide), ET-l (l-21), ET-2 (l-21), and ET-3 (1-21) were purchased from Peptide International (Louisville, KY). [125I]ET-1 was from Amersham. Human umbilical vein endothelial cells and media for their growth were from Clonetics (San Diego, CA). Polyclonal antibody against ET-l was from Biodesign International (Kennebunkport, ME). Phosphoramidon, pepstatin A, and leupeptin were from Boehringer Mannheim. E-64, l,lO-phenanthroline, thiorphan, bestatin, and soy bean trypsin inhibitor were obtained from Sigma. Triton X-100 (hydrogenated) and Tween 20 were from Calbiochem. Ricinus communis agglutinin (RCA)-1 resin was from EY Laboratories, Inc (San Mateo, CA). Monoclonal antibody against neutral endopeptidase 24.11 (NEP 24.11) [also known as common acute lymphoblastic leukemia antigen (CALLA)], J5, was from Coulter (Hialeah, FL). Rabbit kidney NEP 24.11 was obtained from Dr. P. Crine (Department of Biochemistry, University of Montreal, Quebec). Protein A-sepharose was obtained from Pharmacia LKB Biotechnology Inc. Buffers: Buffer A is 60 mM KPi, pH 7.4/10 mM EDTA/8 mM NaN3/0.1% bovine serum albumin/O.l% Tween 20; buffer B is 60 mM KPi, pH 7.4/10 mM EDTA/8 mM NaN3/O.25% gelatin (wt/vol); buffer C is 20 mM Tris-HCl, pH 7.5/0.02% NaN3; buffer D is 10 mM Tris-HCl, pH 7.5/0.25 M sucrose/20 mM KCl; buffer E is 50 mM Tris-HCl, pH 7.2150 mM NaCl/0.02% Protease inhibitors cocktail is 1 mM NaN3/0.2% Triton X-100 (hydrogenated); phenylmethylsulfonyl fluoride/O.05 mM pepstatin A/O. 1 mM leupeptin. Measurement of Endothelin-1 Converting Enzyme Activity: The assay measured the production of ET- 1 essentially as described with minor modifications (15). The reaction mixture (50 l.tl) contained 10 l.tM big ET-l, 100 mM Hepes-KOH (pH 7.0), 0.25% Triton X-100, 0.01% NaN3,0.2 mM phenylmethlysulfonyl fluoride, 0.02 mM pepstatin A, 0.1 mM leupeptin, and the enzyme as indicated. After incubation for the indicated period stopped by the addition of EDTA to give a final concentration diluted with buffer A and the generated ET-l was measured RIA mixture (250 ~1) contained the antibody against ET-l, (15000 cpm) in buffer A. The order of the addition was ET-
of time at 37 OC, the reaction was of 10 r&I. This final mixture was by radioimmunoassay (RIA). The an ET-l sample, and [ l251] ET-l 1 sample, followed by antibody, and
Vol. 56, No. 26, 1995
Endothelin Converting Enzyme
2333
then [125I]ET-1.
After incubation at 4 oC for 16 hr, unbound ET-l was co-precipitated by the addition of charcoal (2.4%, wt/vol)/dextran (0.24%, wt/vol) suspension (125 p.1) in buffer B. The amount of ir-ET-l was measured by counting the supernatant and determined from the standard curve. The cross reactivity to big ET-l was less than 0.01% and the detection limit was 1 fmol. One unit of enzyme is defined as the amount generating 1 pmol of immunoreactive (ir) ET-l per min. Cell Culture: BAEC and HUVEC were grown as described (15). The typical enzyme preparation was performed after obtaining confluent cells at 5 to 8 passages from 50-100 roller bottles. The membrane fractions were prepared essentially the same as before (15). A permanent human cell line (EA.hy926), derived by fusing human umbilical vein endothelial cells with the permanent human cell line A549 (derived from a human lung carcinoma), was obtained from Dr. C.-J. S. Edge11 (Department of Pathology, University of North Carolina at Chapel Hill, NC) and was grown as described (19). Immunoprecipitation: The mixture (270 pl) contained protein A-sepharose (70 l,tl) which was equilibrated with 50 mM Hepes-KOH (pH 7.4), and J5 (50 pg), a monoclonal antibody against neutral endopeptidase 24.11 (NEP 24.11). After incubation for 20 h at 4 oC, the mixture was centrifuged for 5 min at 10,000 x g. The pellet was then washed five times with 50 mM HepesKOH (pH 7.4) containing 3% BSA and incubated for 1 day at 4 oC either with RCA-I purified enzyme (23.7 pg) from EA.hy926 or rabbit kidney NEP 24.11 (1.7 pg) in buffer C containing 0.5% Triton X-100 (hydrogenated) and protease inhibitors cocktail (total volume 500 l.~l). After centrifugation as above, the supernatant (36.5 pl) was assayed for the enzyme activity as described above. Partial Purification of ECE from EA.hy926: All operations were carried out at O-4 oC unless otherwise noted. The cells in each roller bottle were washed with phosphate-buffered saline and gently scraped. These cells were washed further with phosphate-buffered saline followed by buffer D and frozen immediately in liquid nitrogen. The cells from 150 roller bottles were suspended in 100 ml of buffer D containing protease inhibitors cocktail and were homogenized via nitrogen cavitation (600 psi, 10 min.) and centrifuged at 5,000 x g for 20 min. This process was repeated with the pellet resuspended in 100 ml of buffer D containing protease inhibitors cocktail. The combined supernatant was then centrifuged at 20,000 x g for 35 min. The resulting supematant was further centrifuged at 100,000 x g for 1 h. The pellet was washed with 120 ml of buffer C, resuspended in 40 ml of buffer C containing 0.5% Triton X-100 (hydrogenated) and protease inhibitors cocktail, and was stirred gently for 1 h (membrane fraction). The clear supematant was obtained by centrifugation at 100,000 x g for 1 h (detergent extract). For RCA-I, the detergent extract was applied at a flow rate of 0.15 ml/mm onto a 4-ml RCA-I column (0.5 x 20 cm) equilibrated with buffer E including protease inhibitors cocktail. The column was washed with the equilibration buffer until A280 nm of the eluate was less than 0.03, and the activity was eluted with buffer E containing 0.5 M galactose at a flow rate of 0.15 ml/min. Fractions of 4 ml were collected and peak fractions were pooled (RCA-I fraction, 20 ml).
Results EA.hy926 Possesses Two Endothelin-1 Converting Activities. When endothelin-1 converting activity from the membrane fraction was tested in the presence of phosphoramidon across a wide range of concentrations (0.1 nM - 60 /.tM), a biphasic curve was produced, consistent with the inhibition of two distinct enzyme species with two different affinities for phosphoramidon (Fig. 1). The resulting data set could be fit to two separate IC50 curves,
Endothelin
2334
Converting Enzyme
Vol. 56, No. 26, 1995
giving IC50 values of 2.5 f 0.45 nM (activity I) and 1.5 k 0.48 pM (activity II). Detergent extraction followed by RCA-I chromatography gave 38.5- and 32.4-fold purification for total endothelin-1 converting activity and activity II (Table I). However, as indicated by the biphasic curve in the presence of phosphoramidon, this procedure did not separate the two activities (data not shown).
2.5
-
1.5
-
1
-
= E e 7 iTI C
0.5
-
lo-‘*
IO.6
10-E
lo-‘0
[Phosphoramidon]
1 o-4
(M)
Fig. 1 Presence of two endothelin-1 converting activities in EA.hy926. The membrane fraction (5.9 pg) from EA.hy926 was incubated at 37 oC for 1 hr with the indicated phosphoramidon concentrations. The amount of ET- 1 generated was analyzed by RIA.
Table I Partial Purification
of ECE from EA.hy926a Total Activity
Fraction
Total Volume
ml
Membrane Detergent extract RCA-I
Protein Concentration
Absorbance(280
I+II
nm)
IIb
units
Specific Activity 1+11
nb
unitdmVA2gO
nm
40.0 39.4
0.902c 0.387c
1324 1159
382 301
3.7 7.6
1.1 2.0
20.0
0.386
1101
275
142.6
35.6
aECE was partially purified from EA.hy926 grown in 150 roller bottles to con fluency. bActivity was measured in the presence of 100 ~_LM thiorphan, which completely inhibited activity I (see “Results”). cAbsorbance shown was measured after lo-fold dilution.
Vol. 56, No. X,1995
EndothelinConvertingEnzyme
lo-8 [Thlorphan]
2335
1 o-7 (M)
Fig. 2 Inhibition of activity I by thiorphan. The detergent-extracted fraction (2.2 pg) from EA.hy926 was incubated at 37 OC for 1 hr with the indicated thiorphan concentrations. The amount of ET- 1 generated was analyzed by RIA. Activity I, Which Has an Affinity For Phosphoramidon In the nM Range, Is Due To Neutral Endopeptidase 24.11 (NEP 24.11). It has been suggested that NEP 24.11 cleaves big ET- 1 at several sites including at the Trp-2 1-Val22 bond (20). Vijayaraghavan et al. also reported that NEP 24.11 can efficiently hydrolyze ET-l with a kcat value of 13 1.4 mitt-1 (2 1). Recently Murphy et al. reported the relative half-lives for hydrolysis of ET-I, big ET-I, and C-terminal fragment (22-38) by NEP 24.11 (22). When the inhibition of endothelin-1 converting activity from the detergent extract was tested in the presence of thiorphan, a NEP 24.11 inhibitor with a Ki value of 2.2 nM (23), an IC50 value of 42.4 + 10.9 nM was obtained (Fig. 2). Also, immunoprecipitation of the RCA-I fraction with J5, a monoclonal antibody against human NEP 24.11, showed 80 % inhibition of endothelin- 1 converting activity. In contrast, rabbit kidney NEP 24.11 did not coprecipitate with J5 (Table II). These data, including IC50 values of 2.5 and 42.4 nM for phosphoramidon and thiorphan, respectively, and coprecipitation of the activity with a monoclonal antibody against human NEP 24.11, indicate that the activity I discussed above is due to NEP 24.11. Activity II, Which Has an Affinity For Phosphoramidon In the p.M Range, Is Similar To ECE Activities From HUVEC And BAEC. In order to test the inhibition of activity II from EA.hy926 by thiorphan, the enzyme reaction was performed in the presence of 100 nM phosphoramidon. Under these conditions, activity I should be completely inhibited (Fig. 1). Thiorphan, at concentrations up to 100 pM, did not inhibit activity II. When phosphoramidon inhibition of activity II was measured in the presence of 100 pM thiorphan in order to completely inhibit activity I, an IC50 value of 1.8 + 0.21 p.M was obtained (Fig 3). This value is comparable to those from the primary cell lines, HUVEC (2.7 pM) and BAEC (1.8 pM) (15). Other characteristics of activity II which are similar to ECE activities from HUVEC and BAEC include the following: Activity II was abolished by phosphoramidon, EDTA, and l,lO-phenanthroline but was not inhibited by pepstatin A, phenylmethlysulfonyl fluoride, soybean trypsin inhibitor, leupeptin, or E-64 - consistent with
Endothelin
2336
Converting
Vol. 56, No. 26,1995
Enzyme
Inhibitors of cathepsin E (SQ 32056) and angiotensin characteristics of a metalloprotease. converting enzyme (captopril and enalaprilat) did not inhibit activity II. Bestatin, an inhibitor for various aminopeptidases, did not affect the activity (Table III). The pH optimum was also narrow with a maximum at neutral pH (data not shown). Additionally, the specific ECE activity of 5.4 units/ml/mg membrane fraction from EA.hy926 (Table I) was comparable to that from primary HUVEC (3.9 units/ml/mg membrane fraction).
Table II Precipitation
of NEP 24.11 Activity by a Monoclonal Endothelin-1
Enzyme
Antibody, J5a
Converting Activity
(% of Control) RCA-I fraction from EA.hy926
19.9 f 2.7
Rabbit kidney NEP 24.11
90.5 + 8.5
aIn each experiment, the indicated enzyme was incubated for 1 day with J5 bound to protein A-sepharose as described in “Materials and Methods.” After centrifugation, the enzyme activity remaining in the supernatant was measured. All experiments were carried out in duplicates. The average activity is expressed as a percentage of the control (enzyme incubated with protein A-sepharose alone).
l -x.
.
-
Fig. 3 Inhibition of ECE activity (activity II) by phosphoramidon. The RCA-I fraction (1.9 pg) from EA.hy926 was incubated in the presence of 100 PM thiorphan at 37 oC for 1 hr with the indicated phosphoramidon concentrations. The amount of ET-l generated was analyzed by RIA.
Vol. 56, No. 26, 1995
Endothelin
Converting
Enzyme
2337
Table III Inhibition
Profile of ECE Activity from EA.hy926 Concentration
Inhibitors
ECE Activity a (% of Control)
Phosphoramidon
50 pM
EDTA
50 pM
1, 10-Phenanthroline
100 pM
73.6
500 pM
4.0
6.3 6.4
100 pM
112.6
500 pM
8.6
Captopril
100 pM
97.9
Enalaprilat
100 pM
121.0
SQ 32056
100 pM
98.7
200 pM
130.7
Soybean trypsin inhibitor
0.13 mg/ml
150.4
Pepstatin A
20 pM
136.2
Leupeptin
200 pM
135.7
E-64
500 pM
140.0
Bestatin
150 pM
95.8
EGTA (2 mM)
Phenylmethlysulfonyl
fluoride
aThe RCA-I fraction (2.6 pg) of EA.hy926 was incubated in the presence of 100 pM thiorphan at 37 oC for 1 hr with or without the indicated protease inhibitors. The amount of ET-l generated was analyzed by RIA. The amount of ir-ET-l in the control was 7.4 pmol. The final concentration of organic solvent (dimethyl sulfoxide, isopropyl alcohol, ethanol, or methanol) was 1%.
For comparison, endothelin-1 converting activity of the membrane fraction from HUVEC was tested in the presence of thiorphan (0.01-100 PM). These conditions produced approximately IO20 % inhibition compared to the control (without thiorphan), suggesting that at most lo-20 % of the total endothelin-1 converting activity is due to NEP 24.1 I in HUVEC (data not shown). The combined results indicate that the level of NEP 24.11 activity in HUVEC is much lower than that in the hybrid cells, EA.hy926. The expression of NEP 24.11 in HUVEC has been reported (24). Z& Inhibits ECE Activity. When additional ZnC12 was added to the assay, the ECE activity
was inhibited
in a dose-
dependent manner with an IC50 value of 3.2 * 0.2 pM (data not shown). Similar Zn2+ sensitivity was also observed with the ECE activity from HUVEC (IC50: 10.8 + 0.8 pM) (data not shown). At 10 pM, neither M&12, MgC12, CaC12, NiC12, nor CoC12 had any effect.
Endothelin Converting Enzyme
2338
Vol. 56, No. 26, 1995
Substrate Specificity. ECE showed a high specificity for big ET- 1. When ECE was incubated with big ET-l, big ET-2, or big ET-3, it showed a specificity ratio of 40:2.5:1 for big ET-1:big ET-2:big ET-3 (Fig. 4). Similar specificity was also observed with the ECE activity from HUVEC (data not shown). Sawamura et al. reported similar substrate specificity for their thiorphan-insensitive ECE activity from porcine lung membrane (20). 1.2
1
2 Time
3
(hr)
Fig. 4 Substrate specificity of ECE. The RCA-l fraction (2.3 l.tg) from EA.hy926 was incubated with 1 pM big ET-l (a), big ET-2 (0), or big ET-3 (A) in the presence of 100 pM thiorphan for the indicated period of time. The product produced was quantified by RIA from standard curves generated using known amounts of ET-l , ET-2, or ET-3. Discussion
Several groups have reported phosphoramidon-inhibitable ECE activities from various sources (13-15,20,25-29). It is noteworthy that the majority of these ECE activities gave IC50 values in the low pM range for phosphoramidon regardless of the source (14-15,20,27-29). One can use this level of phosphoramidon sensitivity (ie: PM range) as a tool to differentiate genuine ECE activity from other converting activities, assuming that this is a defining property of ECE. Other e.g., angiotensin converting enzyme, NEP 24.11, phosphoramidon-sensitive enzymes, carboxypeptidase A, P. aeruginosa elastase, and thermolysin, are known to have much tighter binding for phosphoramidon (in the nM range of IC50) (30). We therefore used this characteristic of phosphoramidon sensitivity in the pM range to search for ECE activity from the transformed human endothelial cell line, EA.hy926, and showed that there are two endothelin-1 converting activities based on the biphasic curve. It was demonstrated that the endothelin-1 converting activity with an IC50 value of 2.5 nM for phosphoramidon is due to the presence of NEP 24.11. As shown in Fig. 2, when thiorphan inhibition of endothelin-1 converting activity from the detergent extract of EA.hy926 was tested, a maximum inhibition level of approximately 75-80% was reached at about 1 l.tM and remained at
Vol. 56, No. X,1995
Endothelin Converting Enzyme
2339
this level even up to 100 PM (Fig. 2). This same level of inhibition was seen for the first curve in the biphasic phosphoramidon titration (Fig. 1) and when RCA-I purified ECE from EA.hy926 was immunoprecipitated with a monoclonal antibody against human NEP 24.11 (Table II). These results indicate that ECE activity is approximately 20-25% of the total endothelin-1 converting activity from EA.hy926. While this paper was in preparation, the use of this cell line for studying ECE was also described by Waxman et al. (3 1). They reported that the human lung carcinoma A549 cells contain less than 10 % of the ECE activity present in EA.hy926 cells. Since the ECE activity from the primary HUVEC is 80-90 % of total endothelin-1 converting activity and is comparable to that from the hybrid cells, the overexpression of NEP 24.11 in EA.hy926 cells was probably a consequence of the hybridization procedure. Others have previously suggested that NEP 24.11 is expressed in EA.hy926 (32, 33). The current study indicates that the ECE activity of HUVEC has been preserved in EA.hy926. By completely inhibiting the NEP 24.11 activity in EA.hy926 using thiorphan in a manner that does not affect the ECE activity, the ECE activity has been successfully characterized based on pH dependence, protease inhibitor effect, and substrate specificity. We also demonstrated that the ECE activity is inhibited by additional Zn2+ with 1C50 values of 31 1 l.tM. Waxman et al. reported the effect of ZnC12 on EDTA-treated and completely inactivated enzyme. The maximal ECE activity was obtained at 50 l.tM ZnC12 but a higher concentration (100 PM) was inhibitory (31). This difference in the inhibitory ranges of ZnC12 is likely due to the different residual ZnC12 concentrations in partially purified enzymes. High concentrations of Zn2+ (in the mM range) are known to often inhibit metalloproteases due to the formation of zinc monohydroxide that bridges the catalytic zinc ion to a side chain in the active site of the enzyme (34). Besides metalloproteases there are some cases where Zn 2+ inhibition has been observed. The cysteine protease, calpain, and the acid protease, HIV protease, are both also known to be inhibited by Zn2+ in the PM range (Wang, K., personal communication) (35). Further study might be useful to elucidate the mechanism of this Zn2+ inhibition of ECE activity. It has been shown here that the specific ECE activity obtained from EA.hy926 is in the same range as that from primary HUVEC (See “Results”). Therefore one can utilize this cell line for ECE purification without sacrificing the specific activity. This study demonstrates that a permanent human endothelial cell line, EA.hy926, useful for the study of ECE and its regulation of ET-l production.
could be very
Acknowledgments
We thank
Kenneth
Ray III and Gregory
Olds for assistance
in cell
culture
and enzyme
assays.
References
1.
2. 3.
M. YANAGISAWA, H. KURIHARA, S. KIMURA, Y. TOMOBE, M. KOBAYASHI, Y. MITSUI, Y. YAZAKI, K. GOT0 and T. MASAKI, Nature 332 411-415 (1988). T. KASHIWABARA, Y. INAGAKI, H. OHTA, A. IWAMATSU, M. NOMIZU, A. MORITA and K. NISHIKORI, FEBS Lett. 247 73-76 (1989). S. KIMURA, Y. KASUYA, T. SAWAMURA, 0. SHINMI, Y. SUGITA, M. YANAGISAWA, K. GOT0 and T. MASAKI, J. Cardiovasc. Pharmacol. 13 ISUDDI. 5) S5-S8 (1989).
234a
4. 5. 6. 7. 8. 9. 10.
11.
12. 13. 14. 15. 16. 17. 18.
19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
Endothelin Converting Enzyme
Vol. 56, No. 26, 1995
P. D’ORLEANS-JUST, P. S. LIDBURY, T. D. WARNER and J. R. VANE, Biochem. Pharmacol. 39 R21-R22 (1990). G. M. RUBANYI and L. H. BOTELHO, FASEB J. 5 2713-2720 (1991). A. M. DOHERTY, J. Med. Chem. 35 1493-1508 (1992). Y. MATSUMURA, K. HISAKI, M. TAKAOKA and S. MORIMOTO, Eur. J. Pharmacol. 185 103-106 (1990). T. FUKURODA, K. NOGUCHI, S. TSUCHIDA, M. NISHIKIBE, F. IKEMOTO, K. OKADA and M. YANO, Biochem. Biophys. Res. Commun. 172 390-395 (1990). R. IKEGAWA, Y. MATSUMURA, Y. TSUKAHARA, M. TAKAOKA and S. MORIMOTO, Biochem. Biophys. Res. Commun. 171669-675 (1990). T. SAWAMURA, Y. KASUYA, Y. MATSUSHITA, N. SUZUKI, 0. SHINMI, N. KISHI, Y. SUGITA, M. YANAGISAWA, K. GOTO, T. MASAKI and S. KIMURA, Biochem. Biophys. Res. Commun. 174 779-84 (1991). J. E. BIRD, T. L. WALDRON, D. K. LITTLE, M. M. ASAAD, C. R. DORSO, G. DIDONATO and J. A. NORMAN, Biochem. Biophys. Res. Commun. 182 224-231 (1992). D. M. POLLOCK and T. J. OPGENORTH, Am. J. Physiol. &&I R257-R263 (1991). K. OKADA, Y. MIYAZAKI, J. TAKADA, K. MATSUYAMA, T. YAMAKI and M. YANO, Biochem. Biophys. Res. Commun. 171 1192-l 198 (1990). Y. HIOKI, K. OKADA, H. ITO, K. MATSUYAMA and M. YANO, Biochem. Biophys. Res. Commun. 174 446-45 1 (199 1). K. AHN, K. BENINGO, G. OLDS and D. HUPE, Proc. Natl. Acad. Sci. USA. Sq 86068610 (1992). K. SHIMADA, M. TAKAHASHI and K. TANZAWA, J. Biol. Chem. 269 18275-18278 (1994). D. XU, N. EMOTO, A. GIAID, C. SLAUGHTER, S. KAW, D. DEWIT and M. YANAGISAWA, Cell a,473-485 (1994). T. IKURA, T. SAWAMURA, T. SHIRAKI, H. HOSOKAWA, T. KIDO, H. HOSHIKAWA, K. SHIMADA, K. TANZAWA, S. KOBAYASHI, S. MIWA and T. MASAKI, Biochem. Biophys. Res. Commun. 203 1417-1422 (1994). C.-J. S. EDGELL, C. C. MCDONALD and J. B. GRAHAM, Proc. Natl. Acad. Sci. USA. 85) 3734-3737 (1983). T. SAWAMURA, 0. SHINMI, N. KISHI, Y. SUGITA, M. YANAGISAWA, K. GOTO, T. MASAKI and S. KIMURA, Biochim. Biophys. Acta 1161295-302 (1993). J. VIJAYARAGHAVAN, A. G. SCICLI, 0. A. CARRETERO, C. SLAUGHTER, C. MOOMAW and L. HERSH, J. Biol. Chem. 265 14150-14155 (1990). L. J. MURPHY, R. CORDER, A. I. MALLET and A. J. TURNER, Br. J. Pharmacol. 113, 137-142 (1994). A. BEAUMONT, B. BARBE, H. L. MOUAL, G. BOILEAU, P. CRINE, M.-C. FOURNIE-ZALUSKI and B. P. ROQUES, J. Biol. Chem. 267 2138-2141 (1992). C. LLORENS-CORTES, H. HUANG, P. VICART, J.-M. GASC, D. PAULIN and P. CORVOL, J. Biol. Chem. 267 14012-14018 (1992). K. OHNAKA, R. TAKAYANAGI, T. YAMAUCHI, H. OKAZAKI, M. OHASHI, F. UMEDA and H. NAWATA, Biochem. Biophys. Res. Commun. 168 1128-l 136 (1990). Y. MATSUMURA, Y. TSUKAHARA, K. KUNINOBU, M. TAKAOKA and S. MORIMOTO, FEBS Lett. 305 86-90 (1992). K. OHNAKA, M. NISHIKAWA, R. TAKAYANAGI, M. HAJI and H. NAWATA, Biochem. Biophys. Res. Commun. 185 611-616 (1992). M. TAKAHASHI, Y. MATSUSHITA, Y. IIJIMA and K. TANZAWA, J. Biol. Chem. 268 21394-21398 (1993). K. OHNAKA, R. TAKAYANAGI, M. NISHIKAWA, M. HAJI and H. NAWATA, J. Biol. Chem. 268 26759-26766 (1993). J. C. POWERS and J. W. HARPER, Protease Inhibitors, A. J. Barrett and G. Salvesen (Eds), 219-289, Elsevier, Amsterdam (1986).
Vol. 56, No. 26, 1995
31. 32. 33. 34. 35.
Endothelin Converting Enzyme
L. WAXMAN, K. P. DOSH, S. L. GAUL, S. WANG, R. A. BEDNAR and A. M. STERN, Arch. Biochem. Biophys. 308 240-254 (1994). 0. SAIJONMAA, T. NYMAN, U. HOHENTHAL and F. FYHRQUIST, Biochem. Biophys. Res. Commun. 181529-536 (1991). R. CORDER, V. J. HARRISON, N. KHAN, E. ANGGARD, and J. R. VANE, J. Cardiovasc. Pharmacol. 22 (SUDD~.8) S73-S76 (1993). K. S. LARSEN and D. S. AULD, FASEB J. 2 4754 (1988). Z.-Y. ZHANG, I. M. REARDON, J. 0. HUI, K. L. O’CONNELL, R. POORMAN, TOMASSELLI and R. HEINRIKSON, Biochemistry 30 8717-8721 (1991).
2341
A.