Endothelial cells play a critical role in the circulation and are known to produce and/or release a variety of vasoactive substances (1). Among these are the potent vasorelaxants, such as prostacyclin and endothelium-derived relaxing factor (NO), and hyperpolarizing factor (EDHF), and vasoconstricting peptides, such as angiotensin II and endothelin. None of the endothelium-derived peptides has heretofore been considered to be amidated at the C-termini. Many vasoactive neuropeptides such as substance P (SP), calcitonin gene-related peptide (CGRP) and vasoactive intestinal peptide are amidated, and this final reaction process confers their biologic activity (2,3). It is now clear that carboxy-terminal amidation entails two sequential enzymatic steps. The first enzyme, peptidylglycine α-monooxygenase (PAM) catalyzes formation of the α-hydroxyglycine derivative of the glycine extended precursor. The second enzyme, peptidylamidoglycolate lyase (PGL), catalyzes dealkylation of the α-hydroxyglycine derivatives to produce the final amidated product plus glyoxylate. Recently Oldham et al. (4) discovered that these two enzymes, PAM and PGL, are present in both the cytosol and membrane fractions of cultured bovine aortic endothelial cells (BAECs). Endothelial PAM and PGL exhibit enzymologic characteristics that correspond precisely to those of PAM and PGL from other tissues.
A number of novel substrates and potent inhibitors for PAM and PGL have been developed. Among these are a PAM mechanism-based irreversible inactivator, 4-phenyl-3-butenoic acid (PBA), and a potent diketo competitive inhibitor, 5-acetamido-2,4-diketo-6-phenylhexanoic acid and the prodrug methyl ester of this diketo inhibitor (diketo ester), which blunt the actions of PGL (4-8). This enzymatic pathway and sites for inhibitors are shown in Fig. 1.
We recently reported that the glycine extended precursor of SP (SP-Gly) causes relaxation of rat aortic strips with the same potency and efficacy as SP itself, and that PBA caused marked inhibition of the relaxation actions of SP-Gly but not those of SP (9). We also found that both PAM and PGL can be secreted by endothelial cells. Furthermore, inhibition of PAM also decreased basal NO production by cultured cells. These findings suggested a role for amidative processing of peptides by endothelial cells in local vascular regulation.
In this study, we further investigated this hypothesis using both in vitro and in vivo models. We determined and compared the effects of inhibiting PAM or PGL on vasorelaxation to SP and SP-Gly to determine the effect of these inhibitors on the basal vascular tone.
MATERIALS AND METHODS
Male Sprague-Dawley rats (250-350 g) were killed by decapitation. The thoracic aorta of each rat was rapidly removed, cleaned from the adjacent tissue, and cut into four ring segments of 4 mm in length. Two metal hooks were carefully passed through the lumen of each ring and then mounted in 25-ml organ bath containing Krebs solution under 2 × 10−2 Newtons of tension. Krebs solution had the following composition (mM): NaCl, 118; KCl, 4.75; CaCl2, 2.54; MgSO47H2O, 1.2; KH2PO4, 1.19; NaHCO3, 23; and glucose, 11; the pH of the solution was 7.35-7.45. Rings were equilibrated for 60 min, with solution changes every 15 min. The bathing solution was kept at 37°C and was continuously aerated with a mixture of 5% CO2 and 95% O2. For experiments, vessels were preconstricted with 0.3 μM 1-phenylephrine (PE). This concentration was chosen by constructing a concentration-response curve to PE, which showed that this concentration produced 86 ± 3% of the maximal response to PE. Vessels with good integrity of the endothelial cells, as assessed by ∼80% relaxation to 0.1 μM acetylcholine, were used.
To establish the role of NO in SP- or SP-Gly-induced relaxation, concentration-response curves to these agents were constructed in the absence and the presence of L-nitroarginine methyl ester (L-NAME; 0.6 M) or indomethacin (60 μM). Indomethacin or L-NAME was added 30 min before the construction of concentration-response curves. In other experiments, we determined the dependence of the vasorelaxation responses on endothelium. The endothelium of the vascular ring was removed by gently rolling the luminal surface of the vessel segments over a small metal rod. Complete denudation of the endothelium was confirmed by the lack of any relaxant responses to acetylcholine. All relaxant responses were expressed as a percentage of the maximal relaxation response produced by acetylcholine.
Cumulative concentration-response curves to either SP or SP-Gly in the absence and presence of an amidation inhibitor were constructed over a period of 1 h. For each experiment, one of the four vessel segments of a rat was used as a time control for PE contraction. The other three segments were randomly assigned for treatments. Rings were pretreated with the 10 μM PBA, the irreversible mechanism-based inhibitor of PAM and the reversible diketo inhibitor of PGL or diketo ester for 30 min. After this incubation, fresh solution containing 10 μM inhibitor was added to baths. Concentration-response curves were then constructed, and tissues were washed extensively with buffer lacking diketo inhibitor, and other concentration-response curves were constructed. SP or SP-Gly was allowed to act for 10 min, and the degree of relaxation was monitored. In some experiments, concentration-response curves for acetylcholine were constructed before and after 30-min incubation with 10 μM amidation inhibitor.
Direct measurement of nitric oxide production in cultured endothelial cells
NO production was measured directly with a NO meter (MARK II ISO-NO; World Precision Instruments, Sarasota, FL, U.S.A.) connected to a polargraphic NO electrode as previously described (10). BAECs were isolated and amplified using methods described previously (11). In brief, BAECs (passages four through six) were plated on 24-well plates and grown to confluence in M 199 medium. Just before NO measurement, cells were washed twice with phosphate buffer and then bathed in fresh medium. Next, the NO sensor probe was inserted vertically into the wells such that the tip of the electrode was submerged 2 mm under the surface of the medium. The electrode was routinely calibrated with graded concentrations of the NO donor S-nitroso-N-acetylpenicillamine (SNAP) from 2 to 250 nM to obtain a standard curve (12).
The wells containing the confluent BAECs were randomly divided into the following treatment groups: SP (1 μM), SP-Gly (1 μM), and acetylcholine (1 μM). A similar set of experiments was performed after 30 min of pretreatment with PBA, diketo inhibitor, or diketo-ester (1 or 10 μM). The peak reading of the meter was recorded, and the amount of NO was calculated from the SNAP standard curve. The basal NO release in nontreated cultures or cultures pretreated with amidation inhibitors also were measured.
Effect of PBA on the peripheral vascular resistance
Male albino New Zealand rabbits weighing ∼3-3.5 kg were anesthetized by intramuscular administration of ketamine (15 mg/kg) and xylazine (30 mg/kg). Animals were ventilated using a Harvard respirator (model 683, Quincy, MA, U.S.A.) at a rate of 35 stroke/min and a tidal volume 25 ml/stroke. Anesthesia was maintained by inhalation of 2% isoflurane in 100% oxygen. An incision was made in the left femoral triangle to isolate the left femoral artery and vein. A heparinized saline-filled PE-50 catheter was introduced into the femoral artery. This catheter was connected to a Statham P23AC pressure transducer (Hato Rey, Puerto Rico) for a continuous blood pressure monitoring on a Grass 7D Polygraph (Grass Instruments, South Natick, MA, U.S.A.). Another PE-50 tube was placed in the femoral vein and used for drug administration.
A branch of the right femoral artery was carefully isolated, and a PE-10 catheter was introduced and secured in place to administer PBA. The main right femoral artery was isolated, and a Doppler blood flow probe (2 SB; Transonic Corp., Ithaca, NY, U.S.A.) was placed for monitoring femoral blood flow. Lead II ECG was recorded and used to trigger a cardiotachometer for heart rate recording. Animals were divided into two groups. In one group (n = 6) and based on the femoral rate of blood flow, rabbits were infused with PBA to produce intraarterial concentrations of 2.4, 24, and 240 μM at a volume rate of 250 μl/min. Each dose was administered for 20 min followed by an observation period of 30 min before starting the subsequent infusion. At the end of PBA administration, animals were observed for 90 min. A second group of rabbits (n = 6) received only saline.
Chemicals and drugs
Substance P, acetylcholine, phenylephrine, M199 medium, indomethacin, nitro-L-arginine methyl ester, and S-nitroso-N-acetylpenicillamine were purchased from Sigma Chemical Co., St. Louis, MO, U.S.A. 4-Phenyl-3-butenoic acid, 5-acetamido-2,4-diketo-6-phenyl-hexanoic acid, and the prodrug, methyl ester form of this diketo inhibitor, as well as SP-Gly, were synthesized as previously described (5,13).
Peripheral vascular resistance of the hindquarter was calculated by dividing the blood pressure by femoral blood flow. The data are expressed as mean values ± SEM. Differences among the means were evaluated using Student's t test. One-way analysis of variance was used to determine if differences exist among treatment groups. The Neuman-Keul's post hoc test also was used. A probability value <0.05 was considered statistically significant.
As we have shown previously, both SP and SP-Gly were essentially equipotent in producing concentration-relaxation curves. The mean percentage relaxations to SP and SP-Gly at 10 μM were 48 ± 5 and 49 ± 4, respectively, compared with the maximal response observed for acetylcholine (10 μM). Relaxation responses to the two peptides occurred with the same latency (∼1 min). To test the role of the vascular endothelium in vasorelaxation responses to SP-Gly, additional experiments were performed using endothelium-denuded vessels. These experiments showed that these responses were completely eliminated by removal of vascular endothelial cells (Fig. 2A). Moreover, the NO synthase inhibitor L-NAME (0.6 mM) also reduced the vasorelaxation responses significantly but not completely (∼35%; Fig. 2B). Inhibition of prostanoid synthesis by treatment with indomethacin (60 μM) had no significant effect on vasorelaxation responses to either peptide (data not shown).
To determine the effect of the amidation inhibitors, we compared the pooled maximal relaxation responses to SP or SP-Gly (0.1-10 μM) without inhibitors (100% of control response for each) with the responses obtained after inhibitor treatment (10 μM;Fig. 3). The pretreatment of vascular strips with PBA slightly blunted their vasorelaxation response to SP, but the difference was not statistically significant. In contrast to this effect on SP, PBA pretreatment markedly reduced vasorelaxation responses to SP-Gly, averaging 74% less. In both instances, the PBA had been washed from the rings before SP and SP-Gly were administered.
The diketo inhibitor did not significantly alter the vasorelaxing response to SP when it was present in the bath or after it was washed out. However, responses to SP-Gly with the diketo inhibitor in the bath were reduced by ∼65%. The response of the rings that had been treated with diketo inhibitor and then washed was not significantly different from that of the untreated controls.
Because the diketo-ester was insoluble in water, ethanol (0.04% final bath concentration) was used as the solvent and examined as vehicle. Compared with vehicle, the diketo-ester did not significantly influence the relaxation responses to SP. The relaxation response to SP-Gly with the diketo-ester in the bath in comparison to vehicle was markedly suppressed by 71%. However, the responses of the rings pretreated with diketo-ester and then washed were not significantly different from those of the untreated controls. None of the amidation inhibitors alone altered basal vascular tension. Furthermore, the presence of 10 μM amidation inhibitors did not alter the vasoconstrictor or vasodilating responses to phenylephrine or acetylcholine, respectively (data not shown). This was evident by the superimposed concentration-response curves to these agents in the absence and the presence of an amidation inhibitor.
Nitric oxide production
Because the PAM/PGL system is present and operative in the vascular endothelial cells and SP and other neuropeptides are NOS agonists, we investigated the actions of these amidation inhibitors on NOS function in producing NO. All three amidation inhibitors caused a reduction in basal unstimulated NOS activity in the cultured BAECs. When tested at a concentration of 10 μM, PBA, diketo inhibitor, and diketo-ester resulted in a reduction in basal NOS activity of 40, 34, and 45%, respectively. These changes were significantly different (Fig. 4). The diketo inhibitor and diketo-ester were also effective at 1 μM in reducing basal NOS activity by 27 and 30%, respectively.
We also examined the inhibitors' effect on NO-stimulated production by BAECs treated with SP and SP-Gly. Both SP and SP-Gly caused similar and significant (p < 0.05) increases in NO production, which did not differ from one another for a particular experimental series (control responses in Figs. 5-7). These responses to the peptides ranged from 23 to 46 nM increases above the basal NO production, whose levels ranged between 42 and 76 nM in the various experimental series. The time to peak production was the same for both agents (∼10 min). PBA at 1 and 10 μM had no significant effect on levels of NO production induced by SP. However, PBA did decrease the responses to SP-Gly by 33 and 63%, at 1 and 10 μM, respectively (Fig. 5). Neither concentration of diketo inhibitor tested had a significant effect on SP-stimulated NO release. However, the diketo inhibitor reduced SP-Gly stimulated NO release by 42% at concentration of 1 μM and by 69% at 10 μM(Fig. 6). Similarly, neither concentration of diketo-ester altered SP-induced NO production in BAECs, but both significantly reduced the SP-Gly response by 40% at 1 μM and by 68% at 10 μM(Fig. 7).
Basal (Fig. 4) and peptide-stimulated NO levels (Fig. 7) were lower in the experimental series that examined the effect of the 10 μM concentration of diketo-ester because these experiments were performed with BAECs from a different isolation. Different BAEC preparations exhibit different basal NOS activities and sensitivities to stimuli.
Further to demonstrate the specificity of the inhibitor's effects, additional control studies were done using acetylcholine in the presence or absence of PBA and the diketo-ester. Neither PBA nor diketo-ester at 10 μM affected acetylcholine-induced NO production (Fig. 8).
Effect of BPA on the peripheral vascular resistance
To test the contribution of amidated peptides in maintaining of the vascular tone and the ability of amidation inhibitors to influence such tone, we examined effects of PBA on the blood flow and vascular resistance of the hindquarter of rabbits. As shown in Fig. 9, intrafemoral artery infusion of PBA at final concentrations of 2.4, 24, and 240 μM in the blood resulted in a dose-dependent increase in the vascular resistance of the hind leg of the rabbit. This effect was more prominent after 60 min after infusion. The infused doses of PBA did not produce noticeable systemic effects. Following the same protocol, control rabbits infused with saline did not show significant changes in their vascular resistance over the experimental period.
Substance P, an amidated neuropeptide found most commonly in sensory nerves, is recognized to affect the cardiovascular function profoundly (14). Work in our laboratory has demonstrated that SP is a vasodilator, depresses cardiac rate and contraction, and may evoke autonomic reflex changes in cardiovascular function (15). Exogenously supplied SP is known to act on cell-surface receptors of endothelial and other cells to cause activation of NOS, which produces NO, an extremely important effector of cardiac and vascular function (16,17). SP has been detected in vascular endothelial cells (18,19); the investigators did not determine whether this SP had been passively accumulated or actively processed within the cells. However, the presence of SP in cultured endothelial cells after several passages strongly suggests that SP is produced within the endothelium. Like other bioactive peptides, SP is generated by posttranslational modifications of a larger precursor peptide. In particular, carboxy terminal α-amidation is essential for the activity of more than half of the known biologically active neuropeptides, including SP (2,3). The enzymes PAM and PGL have been shown to function sequentially in carboxy-terminal amidation of glycine-extended substrates (6,20). Both PAM and PGL are present and secreted in bifunctional and monofunctional forms in cultured BAECs (9). Moreover, the enzymologic characteristics of the endothelial PAM/PGL system are the same as those of amidating enzymes from other tissues. It is thus clear that endothelial cells are capable of producing mature active amidated SP from its glycine-extended precursor. We therefore hypothesized that production of SP by the vascular endothelium plays an important role in determining basal vascular tone as well as in modulating vascular responses during conditions of activation.
To test this hypothesis, we have conducted experiments to investigate the effects of inhibiting PAM or PGL on vasorelaxation responses to SP versus SP-Gly and to determine whether the inhibitors can directly alter basal and stimulated NO production by the vascular endothelium. Finally, we investigated the contribution of the amidated peptides in maintaining vascular tone of the rabbit's hindquarter.
Two classes of inhibitors have been used: an irreversible, mechanism-based inactivator of PAM, referred to as PBA (5,8), and reversible, competitive inhibitors of PGL, referred to as the diketo inhibitor and diketo-ester, its methyl ester proform (13). The methyl ester analogue exhibits enhanced lipid solubility. This likely increases the penetration of the analogue through the plasma membrane; esterase-catalyzed conversion to the active form occurs rapidly in the cytoplasm.
Our data show that SP-Gly, the glycine-extended precursor of SP, causes relaxation of rat aortic strips with the same latency, potency, and efficacy as SP itself. We now demonstrate that the vasorelaxation responses to SP-Gly and SP are endothelial cell dependent, are partially inhibited by the NOS inhibitor L-NAME, and are unaffected by indomethacin. These results provide evidence that the vasorelaxation responses to SP-Gly, as well as SP, are due in part to NO production consequent to NOS activation. This was further confirmed by experiments using an NO electrode to directly measure NO production. The observation that vasorelaxation response to SP and SP-Gly was only partially inhibited (∼35%) by L-NAME may indicate that endothelium-derived NO is only part of the mechanism of vasodilation elicited in rat aorta with SP. Other endothelium-dependent mechanisms may be involved. Further investigation is needed to determine whether this residual vasorelaxation response to SP in these vessels is due to activity of EDHF, as it is in vessels from other species (21,22).
The data reported here also demonstrate that SP-Gly is equally effective as SP in increasing NO release by NOS. It is notable that very specific inhibitors of either PAM or PGL caused marked concentration-dependent inhibition of the effects of SP-Gly on both vasorelaxation and NO release, but did not alter the responses to SP itself. These inhibitors also decreased basal levels of NO release by endothelial cells. NO production in response to SP or acetylcholine is not affected by any of the three amidation inhibitors, nor do the inhibitors alter cell viability, as demonstrated by trypan blue exclusion. Therefore, we believe that the actions of these agents in suppressing basal NO formation could not be a result of cellular toxicity or direct inhibition of NOS, but rather reduction in cellular products of peptide amidation that are NOS agonists. Thus our results demonstrate a link between peptide amidation and NO formation in endothelial cells.
Both diketo inhibitors were more effective than PBA at the lower 1 μM concentration. In studies examining stimulated NO production by SP and SP-Gly, the diketo-ester was the most effective in reducing the response to SP-Gly and had the least effect on responses to SP. Diketo-ester also exhibited the smallest effect on SP-induced vasorelaxation while causing prominent inhibition of SP-Gly-induced relaxation. This difference between diketo-ester and diketo inhibitor may be due to the enhanced penetration of the plasma membrane by the methylated proform diketo-ester. It should also be noted that the actions of PBA lasted longer than those of the other two inhibitors and were not removed by extensive washing. This is fully consistent with the mechanisms of action of these inhibitors.
Interestingly, infusion of PBA into the femoral artery resulted in a significant increase in the peripheral vascular resistance of the hindquarter. This effect was most prominent 1 h after completion of the infusion. This 1-h period is probably required for metabolism of the already available vasodilating amidated peptide(s) before noting the effect of inhibition of de novo synthesis by PBA. It should be noted that inhibition of the maturation of vasoactive peptides in tissues other than endothelial cells may have contributed, to some degree, to increased vascular resistance of the hind limb in the presence of PBA. These data support our hypothesis that amidation products contribute to tonic maintenance of vascular tone and that inhibition of the amidation process significantly alters the vascular resistance. It is worth noting that despite the significant inhibitory effect of amidation inhibitors on NO formation by endothelial cells and an increase in vascular resistance in vivo, these agents had no significant effect on the basal tone of the isolated aortic rings. This seemingly contradicting observation may be due to the differences in sensitivity of the techniques used.
In summary, our data demonstrate that vascular endothelial cells can process the inactive glycine-extended precursor of SP to an active NOS activator and vasorelaxant through a highly functional PAM/PGL system. This indicates that a product of the endothelial cell peptide amidation process can influence NO production. The observation that basal levels of NO production are reduced by inhibition of the PAM/PGL system, in addition to the increase in vascular resistance after intraarterial administration, strongly suggests that an intrinsic peptide amidation product has an important role in the regulation of NOS activity and vascular tone. Moreover, because the induction of NOS activation has been found to have a critical role in mediating the angiogenic and permeability-enhancing effects of exogenous SP (23,24), it is likely that an endothelial cell source of SP could also have a germinal role in regulation of vascular growth and permeability function. The possible mechanisms of release of the formed peptides from endothelial cells are currently under investigation.
Acknowledgment: This research was partially supported by NIH grant GM 40540 and the Medical College of Georgia/Georgia Institute of Technology Biomedical Research and Education Program.
1. Furchgott RF, Vanhoutte PM. Endothelium-derived relaxing and contracting factors. FASEB J
2. Bradbury AF, Smyth DG. Biosynthesis of the C-terminal amide in peptide hormones. Biosci Rep
3. Eipper BA, Mains RE. Peptide α-amidation. Annu Rev Physiol
4. Oldham CD, Li C, Girard PR, Nerem RM, May SW. Peptide amidating enzymes are present in cultured endothelial cells. Biochem Biophys Res Commun
5. Katopodis AG, May SW. Novel substrates and inhibitors of peptidylglycine α-amidating monooxygenase. Biochemistry
6. Katopodis AG, Ping D, Smith CE, May SW. Functional and structural characterization of peptidylamidoglycolate lyase, the enzyme catalyzing the second step in peptide amidation. Biochemistry
7. Ping D, Katopodis AG, May SW. Tandem stereochemistry of peptidylglycine α-monooxygenase and peptidylamidoglycolate lyase, the two enzymes involved in peptide amidation. J Am Chem Soc
8. Bradbury AF, Mistry J, Roos BA, Smyth DG. 4-Phenyl-3-butenoic acid, an in vivo inhibitor of peptidylglycine hydroxylase (peptide amidating enzyme). Eur J Biochem
9. Oldham CD, Li C, Feng J, et al. Amidative peptide processing and vascular function. Am J Physiol
10. Kaesemeyer WH, Caldwell RB, Huang J, Caldwell RW. Pravastatin activates endothelial nitric oxide synthase independent of its lipid lowering actions. J Am Coll Cardiol
11. Behzadian A, Wang X, Jiang B, Caldwell RB. Angiostatic role of astrocytes: suppression of vascular endothelial cell growth by TGF-β and other inhibitory factor(s). GLIA
12. Ichimori K, Ishida H, Fukahori M, Nakazawa H, Murakami E. Practical nitric oxide measurement employing a nitric oxide-selective electrode. Rev Sci Instrum
13. Mounier CE, Shi J, Sirimanne SR, et al. Pyruvate-extended amino acid derivatives as highly potent inhibitors of carboxyl-terminal peptide amidation. J Biol Chem
14. Furchgott RF. The role of endothelium in the responses of vascular smooth muscle to drugs. Annu Rev Pharmacol Toxicol
15. Chiao H, Caldwell RW. The role of substance P in myocardial dysfunction during ischemia and reperfusion. Arch Pharmacol
16. Rees DD, Palmer RM, Hodson HF, Moncada S. A specific inhibitor of nitric oxide formation from L-arginine attenuates endothelium-dependent relaxation. Br J Pharmacol
17. Ziche M, Morbidelli L, Parenti A, et al. Substance P increases cGMP levels of coronary endothelial cells. Life Sci
18. Loesch A, Burnstock G. Ultrastructural localisation of serotonin and substance P in vascular endothelial cells of rat femoral and mesenteric arteries. Anat Embryol
19. Milner P, Kirkpatrick KA, Ralevic V, Toothill V, Pearson J, Burnstock G. Endothelial cells cultured from human umbilical vein release ATP, substance P, and acetylcholine in response to increased flow. Proc R Soc Lond
20. Katopodis AG, Ping D, May SW. A novel enzyme from bovine neurointermediate pituitary catalyzes dealkylation of alpha-hydroxyglycine derivatives, thereby functioning sequentially with peptidylglycine alpha-amidating monooxygenase in peptide amidation. Biochemistry
21. Petersson J, Zygmunt PM, Hogestatt ED. Characterization of the potassium channels involved in EDHF-mediated relaxation in cerebral arteries. Br J Pharmacol
22. Shimizu S, Paul RJ. The endothelium-dependent, substance P relaxation of porcine coronary arteries resistant to nitric oxide synthesis inhibition is partially mediated by 4-aminopyridine-sensitive voltage-dependent K+
23. Ziche M, Morbidelli L, Masini E, et al. Nitric oxide mediates angiogenesis in vivo and endothelial cell growth and migration in vitro promoted by substance P. J Clin Invest
24. Ziche M, Parenti A, Ledda F, et al. Nitric oxide promotes proliferation and plasminogen activator production by coronary venular endothelium through endogenous βFGF. Circ Res