Regulation and Differential Expression of Neutral Endopeptidase 24.11 in Human Endothelial Cells
Abstract Neutral endopeptidase 24.11, a membrane-bound metallopeptidase, cleaves and degrades vasoactive peptides such as atrial natriuretic peptide, endothelin, angiotensin I, substance P, and bradykinin. Therefore, the presence of this metallopeptidase may contribute to the regulation of vascular tone and local inflammatory responses in the vascular endothelium and elsewhere. We determined neutral endopeptidase in cultured human endothelial cells from different vascular beds and studied its regulation by protein kinase C. Neutral endopeptidase was detected in all cultured endothelial cell types. Lowest concentrations were measured in human endothelial cells from umbilical veins (360±14 pg/mg protein), followed by pulmonary and coronary arteries; higher concentrations were found in endothelial cells from the cardiac microcirculation (1099±73 pg/mg protein). Neutral endopeptidase content increased during cell growth but was not affected by endothelial cell growth factor or modifications of the growth medium. Stimulation of protein kinase C with 1-oleoyl-2-acetyl-rac-glycerol (0.1 to 1 μmol/L) and phorbol 12-myristate 13-acetate (0.01 to 0.1 μmol/L) induced a time- and concentration-dependent increase of endothelial cells that was inhibited by cycloheximide (5 μmol/L), an inhibitor of protein synthesis. Incubation with phospholipase C (1 μmol/L) and thrombin (10 IU/mL) induced upregulation of neutral endopeptidase, resulting in 158±26% and 150±22% increases, respectively, compared with controls. The thrombin effect was inhibited by calphostin C (1 μmol/L), an inhibitor of protein kinase C. Endothelial neutral endopeptidase is constitutively expressed in endothelial cells from different origins and is inducible by thrombin via activation of the protein kinase C pathway.
The membrane-bound metallopeptidase NEP (EC 18.104.22.168) was initially demonstrated in the renal brush border1 ; it was further detected in various organs and cells, such as lung, kidney, brain, thyroid, intestine, leukocytes, lymphoblastic cells, epithelial cells, vascular smooth muscle cells, and the vascular endothelium.2 3 4 NEP modulates the action of various peptide hormones by degrading them to inactive metabolites or converting them to active mediators.5 6 7 NEP has a high affinity for some potent inflammatory and vasoactive peptides such as substance P, bradykinin, N-formyl-l-methionyl-l-leucyl-l-phenylalanine, atrial natriuretic peptide, and endothelin.6 7 8 9 10 11 In rats NEP converts circulating angiotensin I to its active metabolite angiotensin,1 2 3 4 5 6 7 which is an endothelium-dependent vasodilator.5 12
The regulation of NEP expression is not well characterized at the cellular level. In leukocytes and epithelial cells a decrease of NEP activity was observed after treatment with phorbol esters.13 14 An increase in NEP activity was found during the cell growth of transfected epithelial cells and human leukemic cells and after glucocorticoid treatment.15 16 Glucocorticoids also increased NEP gene transcription and enhanced NEP activity in vascular smooth muscle cells of rabbit renal cortex.4 15 The same authors reported an increase of NEP after serum deprivation of cultured smooth muscle cells, indicating a regulatory action of serum factors.
In the present study we investigated the expression of NEP protein and activity during endothelial cell growth and in cultured endothelial cells from different vascular regions, including adult and embryonic phenotypes. Adult endothelial cells were derived from the pulmonary arteries, the microvasculature, and the epicardial coronary arteries of the human heart. Furthermore, we studied the regulation of this peptidase by PKC to elucidate how NEP is regulated in endothelial cells.
HUVECs, HPAECs, and macrovascular HCECs were harvested after collagenase treatment.17 18 The cells were seeded into a 35-mm dish coated with 0.2% gelatin and incubated in culture medium at 37°C and 5% CO2. Microvascular HCECs were isolated as described previously in detail.19 Briefly, capillaries were enzymatically isolated from heart muscle segments and seeded in culture medium. Separation from nonendothelial cells was performed with magnetic tosyl-activated beads (Dynal) coupled to Ulex europaeus I lectin. Purification was 98% endothelial cells determined by flow cytometry (FACS, Becton Dickinson) with a factor VIII–related antigen antibody.19 Culture conditions for all endothelial cells were identical. The culture medium contained medium 199, 20% FCS, 100 IU/mL penicillin, 100 μg/mL streptomycin, 10 mmol/L HEPES, and 0.2 mol/L l-glutamine. ECGF (0.8 μmol/L) was added to the culture medium except in the experiments in which the effect of ECGF on NEP expression was investigated (see below). Macrovascular and microvascular HCECs were cultivated in culture medium containing 20% pooled human serum. All experiments were performed with cells of the second passage.
Measurement of NEP in Endothelial Cells
All experiments were performed with confluent monolayers of endothelial cells grown in 35-mm dishes in 1 mL culture medium. Confluent monolayers were washed twice with phosphate-buffered saline, and incubation medium and drugs then were added. NEP expression at different time intervals (1 to 48 hours) in response to various drugs was studied. The medium was routinely replaced every 24 hours. During observation periods longer than 24 hours, stimulating drugs were not continued in the exchanged medium. Afterwards, supernatants were aspirated, and the cells were washed twice, scraped off the dish with a rubber policeman, and suspended in 1.0 mL Tris buffer (0.1 mol/L, pH 7.4) containing 0.1% Tween 20. The cells were sonicated for 5 seconds (20% output energy) by a Sonifier B-12 (Branson Sonic Power Co) on ice. Samples were immediately frozen and stored at −35°C until further use.
Effect of Culture Medium on Endothelial NEP Expression
HUVECs were cultured under standard conditions until cells were subconfluent. The incubation medium was changed to 1% or 20% FCS containing RPMI medium (instead of medium 199) or standard culture medium containing 20% FCS and medium 199. In one set of experiments subconfluent cells were cultured in medium without ECGF as opposed to standard culture medium containing ECGF (0.8 μmol/L) for 3 days. The medium was replaced every 24 hours. After treatment for 3 days in each set, cells were harvested as described above and NEP was quantified.
ELISA for Measurement of NEP
NEP protein was measured by a sandwich ELISA with the use of three monoclonal antibodies against human NEP, as described previously.20 Recombinant NEP was used as standard reference at concentrations between 40 and 2500 pg/mL. Microtiter plates were coated with the rat anti-NEP monoclonal antibody AL2 for 12 hours; standard dilutions, buffer, or sample was applied in duplicate and incubated for 1 hour. A mixture of two mouse monoclonal antibodies both reacting with human NEP (ALB-1 and MEK 5) was used for the next 1.5-hour incubation phase. After final incubation with alkaline phosphatase–conjugated goat anti-mouse IgG, the substrate (p-nitrophenylphosphate) was added to the plates, and the optical density was determined at 405 nm absorbance (Multiskan MCC/340 MK, Flow Laboratories SA). Intra-assay and interassay coefficients of variation were 3% and 12%, respectively. Protein content in each sample was quantified following the method described by Lowry et al.21
Assay for NEP Activity
The assay for NEP activity has been described in detail by Florentin et al.22 Confluent monolayers of HUVECs were washed with ice-cold 50 mmol Tris-HCl buffer, scraped off the dish with a rubber policeman, and centrifuged at 400g for 4 minutes. The cell pellet of one dish was lysed with 1 mL ice-cold distilled water and immediately frozen for later use. Samples (50 μL) were incubated with 100 μL (50 μmol/L) DAGPNG in the presence of phosphoramidon (0.1 μmol/L) or diluent (total volume, 190 μL) at pH 7.4 and 37°C. The increase of fluorogenic activity was measured at 562 nm with a fluorescence spectrometer (Pharmacia) over 6 hours. Enzymatic activity was defined as the amount of substrate (DAGPNG) cleaved at 37°C per minute (nanomoles per liter per minute per milligram protein=milliunit [international unit] per milligram protein) and was calculated from the maximal slope of the degradation curve. The activity inhibited by phosphoramidon was assigned to NEP activity.
Tissue culture plastic ware was obtained from Becton Dickinson. Medium 199, RPMI, phosphate-buffered saline, FCS, trypsin-EDTA solution, HEPES, and antibiotics were from GIBCO. Collagenase was purchased from Roth. ECGF was from Boehringer. Gelatin, phospholipase C (from Clostridium perfringens), OAG, PMA, thrombin, cycloheximide, p-nitrophenylphosphate, bovine serum albumin, dimethylglutaryl acid, Ulex europaeus agglutinin I lectin, DAGPNG, phosphoramidon, and staurosporine were purchased from Sigma Chemical Co. Calphostin C and H-7 [1-(5-isoquinolinesulfonyl)-2-methylpiperazine] were purchased from Calbiochem. Tris, Tween 20, and dimethyl sulfoxide were obtained from Serva AG. Fluorescence-conjugated low-density lipoprotein was from Paesel & Lorei. The antibody against factor VIII–related antigen was from Behring. Recombinant NEP and mouse anti-NEP monoclonal antibody MEK 5 were a generous gift from Khepri Pharmaceuticals. The rat anti-NEP monoclonal antibody AL2 was provided by Dunn. The mouse anti-NEP monoclonal antibody ALB-1 and alkaline phosphatase–conjugated goat anti-mouse IgG were purchased from Dianova. Calphostin C, cycloheximide, and PMA were solubilized in dimethyl sulfoxide (stock solution, 10 or 1 mmol/L) and diluted in medium 199. Phospholipase C was solubilized in 10 mmol/L dimethylglutaryl acid, 0.1% bovine serum albumin, and distilled water at pH 7.3.
Calculations and Statistics
Each experiment was done with triplicate dishes. Results are shown as mean±SEM (n=number of experiments). Cell numbers were separately counted in each experiment in one representative dish. Statistical significance was analyzed by one-way ANOVA and the Mann-Whitney U test for unpaired samples as appropriate. Probability values less than .05 were accepted as statistically significant.
NEP Expression in Cultured Endothelial Cells Derived From Different Blood Vessels
The results of NEP measurement are summarized in Fig 1⇓. Lowest NEP concentrations were found in HUVECs. Significantly higher values for NEP were found in HCECs and HPAECs. Microvascular HCECs contained threefold higher concentrations of NEP than HUVECs (P<.01, Fig 1⇓).
Effect of Growth and Culture Conditions on Endothelial NEP Expression
Fig 2⇓ summarizes the results of NEP protein and NEP activity measurements in growing HUVECs from days 3 to 6 after the second passage. At day 3 cells were in a subconfluent state, reaching confluence between days 4 and 5. Enzymatic activity, measured by the DAGPNG assay,22 and NEP protein showed a parallel increase between days 3 and 5 after seeding. At day 6 no further increase in NEP protein and activity was observed (Fig 2⇓).
The growth of subconfluent HUVECs in different media (1% or 20% FCS containing RPMI medium or 20% FCS containing medium 199) for 3 days did not affect cellular NEP protein in endothelial cells (RPMI+1% FCS, 376±24 pg/mg protein; RPMI+20% FCS, 341±21 pg/mg protein; growth medium+20% FCS, 433±58 pg/mg protein [n=6]). These treatments did not affect the protein content per dish. ECGF did not affect the NEP content of the cells. Endothelial cells cultured with 10 ng/mL ECGF contained 378±49 pg/mg protein, and those grown in ECGF-free medium (for 3 days) contained 433±29 pg/mg protein (n=9). In contrast, the total protein content per dish was significantly reduced from 624±45 μg protein per dish in ECGF-treated endothelial cells to 397±28 μg protein per dish (P<.01, n=9) in cells cultured without ECGF supplement for 3 days.
Regulation of NEP Expression by PKC Activation
Incubation of confluent HUVEC monolayers with the PKC activator PMA resulted in a time-dependent increase of cellular NEP concentrations, beginning after 4 hours and reaching a maximum at 48 hours (Fig 3⇓). Maximal stimulation of NEP expression was found after 0.1 μmol/L PMA for 48 hours, resulting in a 123±13% increase. Combined incubation of endothelial cells with PMA and cycloheximide, an inhibitor of protein synthesis, inhibited the PMA-mediated increase of NEP (Table⇓). OAG, an analogue of diacylglycerol, significantly increased NEP at concentrations from 0.1 to 1 μmol/L (Fig 4⇓). The effect of OAG on NEP concentrations in endothelial cells was observed after 6 hours of incubation, reached its maximum after 9 hours, and declined thereafter. Increased NEP concentrations were still detected after 24 hours of incubation (Table⇓). The OAG-induced increase of cellular NEP showed a different time pattern compared with the stimulation of endothelial NEP by PMA. The effect of both PKC activators, PMA and OAG, was blocked in the presence of 1 μmol/L calphostin C or staurosporine (Table⇓).
Another activator of the PKC pathway is thrombin. Incubation with increasing concentrations of thrombin induced a concentration-dependent increase of cellular NEP concentrations after 24 hours of incubation, reaching a maximal effect at 10 IU/mL (Fig 5⇓). When thrombin was incubated in the presence of calphostin C (1 to 10 μmol/L), a specific inhibitor of PKC, the thrombin-mediated effect was abolished (Fig 5⇓). Incubation with 10 μmol/L calphostin C alone slightly affected cellular NEP concentrations (225±32 pg/mg protein, n=14), inducing a decrease of 38±8% compared with control, whereas 1 μmol/L calphostin C (409±43 pg/mg protein, n=10) did not change cellular NEP concentrations but also inhibited the thrombin-mediated increase of cellular NEP. In line with these observations, thrombin (10 IU) and PMA (0.1 μmol/L) also increased NEP expression in macrovascular HCECs after 24 hours of incubation (Fig 6⇓). This effect was abolished in the presence of H-7, a PKC inhibitor. Incubation with varying concentrations of phospholipase C also led to NEP induction in endothelial cells, with a maximal effect observed after incubation with 1 μmol/L (158±26% from control, Table⇑). OAG (0.1 μmol/L, 189±44%), thrombin (20 IU, 150±22%), and phospholipase C induced the strongest upregulation of NEP in all experiments performed in this study.
To investigate the role of calcium in NEP regulation, we incubated HUVECs with the calcium ionophore A23187, which did not increase NEP expression (control, 422±56 pg/mg protein, n=12; 0.1 μmol/L A23187, 427±83 pg/mg protein, n=8; 0.3 μmol/L A23187, 331±66 pg/mg protein, n=8; and 1 μmol/L A23187, 301±38 pg/mg protein, n=8).
Using a sensitive ELISA for human NEP 24.1120 we could measure this metallopeptidase in homogenates of various endothelial cells derived from the human vasculature. Comparable NEP activities have been described in the particulate fraction of cultured endothelial cells from aorta and umbilical veins.23 In the present study the lowest concentration of NEP was found in HUVECs. In HCECs and HPAECs, NEP concentration was slightly higher. The highest NEP concentration was observed in microvascular HCECs, a cell type characterized previously in our laboratory.19 All cultures were grown under identical conditions. They were of high purity and showed a typical common pattern of endothelial cells. Differences between macrovascular and microvascular cells have been described regarding some antigenic properties.19 24 25 Microvascular HCECs contain a twofold to threefold higher NEP concentration than endothelial cells from embryonic and venous origins and 40% more NEP than macrovascular HCECs.
The growth experiments demonstrated that the quantities of NEP protein and NEP activity increase proportionally during cell growth. When endothelial cells reached confluence, the highest amounts of NEP protein and activity were detected. The parallel increase of NEP activity confirmed that measurement by the ELISA method recognizes a functionally active peptidase. The dependence on growth suggests a constitutive expression of NEP in endothelial cells. This could also be demonstrated during decelerated growth in the absence of ECGF, when reduced NEP concentration paralleled reduced protein content of the cultures. Indeed, the reduction in the FCS content of the growth medium did not affect the cellular NEP content. Dussaule et al4 reported that low serum concentrations in the culture medium increased NEP expression in rabbit smooth muscle cells. This could not be confirmed in subconfluent HUVECs.
Nevertheless, NEP expression in endothelial cells can be regulated by PKC stimulation. High concentrations of thrombin increased NEP expression in a concentration-dependent manner. PKC inhibition by calphostin C abolished the thrombin effect. OAG and PMA, two direct activators of intracellular PKC, also augmented cellular NEP expression, which was completely blocked in the presence of calphostin C, staurosporine, and H-7.
Stimulation of NEP expression could also be achieved by addition of extracellular phospholipase C to confluent HUVEC monolayers. The mechanism by which extracellular phospholipase C penetrates into endothelial cells and activates intracellular sites is still unclear. It has been reported that extracellular addition of phospholipase C increases intracellular diacylglycerol in cultured mammalian cells26 and imitates the effect of other PKC activators, such as phorbol esters, in cultured human endothelial cells.27 A similar observation was made in the present study. Activation of the intracellular phospholipase C pathway leads to the generation of diacylglycerols and to an inositol 1,3,5-triphosphate–mediated increase of intracellular calcium. The calcium ionophore A23187 also augments intracellular calcium concentrations. In the present study A23187 did not affect NEP expression, which suggests that the upregulation of NEP is not mediated by increased calcium influx or intracellular calcium mobilization.
Interestingly, the PKC activators showed a different time pattern of PKC stimulation. OAG, an analogue of diacylglycerol, increased cellular NEP concentrations after 9 to 12 hours, whereas the induction by PMA was slower and reached its maximal effects after 24 to 48 hours. This difference has been reported by other authors.28 29 The action of the physiological activators of PKC, diacylglycerol and its analogues, is transient and restricted because of enzymatic degradation by diacylglycerol kinase or diacylglycerol lipase.30 The increase of NEP induced by a phorbol ester and OAG was abolished by an inhibitor of protein synthesis, indicating an induction of de novo synthesis by PKC activation. Previous studies in human granulocytes and epithelial cells showed an inhibition of NEP activity by phorbol esters.13 14 Thus the regulation of NEP appears to be a cell type–specific event.
The physiological relevance of endothelial NEP remains speculative. NEP is a very active enzyme with high substrate specificity, especially to bradykinin, substance P, and angiotensin I.2 5 6 It is questionable whether upregulation of NEP affects local peptide concentrations and actions. However, the local capacity of this peptidase depends on its own quantity as well as on local substrate concentrations, which are difficult to determine. Nevertheless, NEP is constitutively expressed in all human endothelial cell types investigated, which suggests a systemic as well as a local regulatory role, eg, in the coronary circulation.1 2 3 We have reported that endothelial NEP is mainly located on the membranes of human endothelial cells and contributes to the degradation of extracellular bradykinin, especially when angiotensin-converting enzyme is inhibited.7 Since NEP is a potent inactivator of vasoactive and inflammatory peptides such as the chemotactic peptide N-formyl-l-methionyl-l-leucyl-l-phenylalanine, bradykinin, atrial natriuretic peptide, angiotensin I, endothelin, and tachykinins,6 7 8 10 its expression and upregulation might influence local vasomotor and inflammatory responses in the macrovasculature as well in the microvasculature. Local stimulation of NEP expression by thrombin, which is generated in high concentrations during acute vascular ischemic events, could theoretically limit the local effects of proinflammatory and vasoactive peptides by induction of NEP.
The present study suggests that the distribution of this peptidase varies in different vascular regions and is higher in the microvasculature than in the macrovasculature. Furthermore, the present data indicate that besides a constitutive expression of NEP, which is maximal in confluent endothelial cells, induction of this peptidase is possible via phospholipase C and PKC activation.
Selected Abbreviations and Acronyms
|ECGF||=||endothelial cell growth factor|
|ELISA||=||enzyme-linked immunosorbent assay|
|FCS||=||fetal calf serum|
|HCEC||=||human coronary endothelial cell|
|HPAEC||=||human pulmonary artery endothelial cell|
|HUVEC||=||human umbilical vein endothelial cell|
|PKC||=||protein kinase C|
|PMA||=||phorbol 12-myristate 13-acetate|
Parts of this work will be included in the MD thesis of Petra Koehne at the Free University of Berlin. The authors wish to thank Karen Vetter and Arthur O’Connor for excellent technical assistance.
Reprint requests to Dr Kristof Graf, German Heart Institute Berlin, Department of Internal Medicine/Cardiology and Angiology, Augustenburger Platz 1, 13353 Berlin, FRG.
Presented in part at the scientific sessions of the American Heart Association, 1992 and 1993, and published in abstract form (Circulation. 1992;86[suppl I]:I-275; Circulation. 1993;88[suppl I]:I-315).
- Received November 16, 1994.
- Revision received January 3, 1995.
- Accepted April 19, 1995.
Kerr MA, Kenny AJ. The purification and specificity of a neutral endopeptidase from rabbit kidney brush border. Biochem J. 1974;134:477-488.
Erdös EG, Skidgel RA. Neutral endopeptidase 24.11 (enkephalinase) and related regulators of peptide hormones. FASEB J. 1989;3:145-151.
Dussaule JC, Stefanske A, Béa ML, Ronco P, Ardaillou R. Characterisation of neutral endopeptidase in vascular smooth muscle cells of rabbit renal cortex. Am J Physiol. 1993;264:45-52.
Abassi ZA, Tate JE, Golomb E, Keiser HR. Role of neutral endopeptidase in the metabolism of endothelin. Hypertension. 1992;20:89-95.
Painter R, Dukes R, Sullivan J, Carter R, Erdös E, Johnson A. Function of neutral endopeptidase on the cell membrane of human neutrophils. J Biol Chem. 1988;263:9456-9461.
Stephenson SL, Kenny AJ. The hydrolysis of a human atrial natriuretic peptide by pig kidney microvillar membranes is initiated by endopeptidase 24.11. Biochem J. 1987;243:183-187.
Vijayaraghavan J, Scicli AG, Carretero OA, Slaughter C, Moomaw C, Hersh LB. The hydrolysis of endothelins by neutral endopeptidase 24.11 (enkephalinase). J Biol Chem. 1990;265:14150-14155.
Erdös E, Wagner B, Harbury C, Painter R, Skidgel R, Fa X. Downregulation and inactivation of neutral endopeptidase 24.11 (enkephalinase) in human neutrophils. J Biol Chem. 1989;264:1419-1423.
Werb Z, Clark EJ. Phorbol diesters regulate expression of the membrane neutral endopeptidase (EC 22.214.171.124) in rabbit synovial fibroblasts and mammary epithelial cells. J Biol Chem. 1988;264:9111-9113.
Borson D, Gruenert D. Glucocorticoids induce neutral endopeptidase in transformed human tracheal epithelial cells. Am J Physiol. 1991;280:L83-L89.
Gräfe M, Bossaller C, Graf K, Auch-Schwelk W, Baumgarten C, Hildebrandt A, Fleck E. Effect of angiotensin converting enzyme inhibition on bradykinin metabolism by human vascular endothelial cells. Am J Physiol. 1993;264:H1493-H1497.
Gräfe M, Auch-Schwelk W, Graf K, Terbeek D, Hertel H, Unkelbach M, Hildebrandt A, Fleck E. Isolation and characterisation of macrovascular and microvascular endothelial cells from human hearts. Am J Physiol. 1994;267:H2138-H2148.
Lowry OH, Rosebrough NH, Farr AL, Randall RJ. Protein measurement with Folin phenol reagent. J Biol Chem. 1951;193:265-275.
Llorens-Cortes C, Huang H, Vicart P, Gasc J-M, Paulin D, Corvol P. Identification and characterisation of neutral endopeptidase in endothelial cells from venous or arterial origins. J Biol Chem. 1992;267:14012-14018.
Gordon E, Danielson P, Nguyens T, Winn R. A comparison of primary cultures of rat cerebral microvascular endothelial cells to rat aortic endothelial cells. In Vitro Cell Dev Biol. 1991;27A:312-326.
Kooistra T, Bosma PJ, Toet K, Cohen LH, Griffioen M, van den Berg E, le Clercq L, van Hindsbergh VWM. Role of protein kinase C and cAMP in the regulation of t-TPA, PAI-1, and PDGF mRNA levels in human endothelial cells. Arterioscler Thromb. 1991;11:1042-1052.
Kolesnick RN, Clegg S. 1,2-diacylglycerols, but not phorbol esters, activate a potential inhibitory pathway of protein kinase C in GH3 pituitary cells. J Biol Chem. 1988;263:6534-6537.