Pharmacological characterization of a novel non-AT1, non-AT2 angiotensin binding site identified as neurolysin

Abstract

The discovery of a novel non-AT₁, non-AT₂ binding site for angiotensins in the rodent brain and testis that is unmasked by the organomercurial compound para-chloromercuribenzoic acid (PCMB) has catalyzed efforts to purify and characterize this protein. We recently reported that this protein is neurolysin and now report upon the specificity of this binding site for various neuropeptides. Competition binding assays in rat brain and testis used ¹²⁵I-Sar¹, Ile⁸ angiotensin II (Ang II) as the radioligand in the presence of saturating concentrations of AT₁ and AT₂ receptor antagonists and 100 µM parachloromercuribenzoate. Primary screening of 36 peptides and other compounds at 10 µM concentration revealed seven peptides that inhibited specific binding > 50%: ghrelin, Tyr¹ S36057 (a melanin-concentrating hormone receptor ligand), orphanin FQ and its congeners (Tyr¹ and Tyr¹⁴), Dynorphin A (1–8), and Ang (1–9). The selective neurolysin inhibitor Proline-Isoleucine dipeptide was inactive at 1 mM. These results suggest that the ability of PCMB to unmask high affinity binding of Ang II to neurolysin is a pharmacological effect and that neurolysin may significantly affect the activity of the reninangiotensin system.

Introduction

Research studies have long determined that angiotensin II (Ang II) plays a significant role in the development of cardiovascular disease and hypertension. According to the Centers for Disease Control and Prevention, the leading cause of death in the U.S. in 2010 was heart disease (595,444 deaths) and the fourth leading cause was cerebrovascular disease (129,180 deaths) (http://www.cdc.gov/nchs/data/nvsr/nvsr60/nvsr60_04.pdf, Accessed October 4, 2012). Ang II is a major effector octapeptide, critical for the maintenance of blood pressure and plasma Na⁺ concentration. The primary role of the renin-angiotensin system (RAS) in the body is to maintain blood volume and electrolyte balance. Dysregulation of the RAS causes several pathophysiological states, of which hypertension is most prominent [1].

Dysregulation can result from altered redox state in the bloodstream which can increase the conversion of angiotensinogen to Ang I [2]. Overexpression of renin, as well as prorenin, also increases Ang I formation. In the classic RAS, angiotensin-converting enzyme (ACE) then converts Ang I into Ang II, the major effector peptide of the RAS. Its well-known roles include stimulation and release of aldosterone, leading to retention of sodium and an increase in blood pressure by vascular smooth muscle contraction and increased sympathetic outflow. Additionally, excessive demand on the heart leads to hypertrophy while direct effects of Ang II on the heart cause fibrosis which compromises cardiac output. Stress on vascular cell walls combined with proinflammatory effects of Ang II lead to an increased deposition of plaque on blood vessel walls leading to atherosclerosis [3]. Ang II exerts these effects primarily on the angiotensin II type I (AT₁) subtype receptor. Ang II is further converted to the heptapeptide Ang III by glutamyl aminopeptidase A. Ang III, which also acts upon the AT₁ subtype, is then converted to the hexapeptide Ang IV by the cleavage of the Arg at its N-terminal by membrane alanyl aminopeptidase N [4]. Ang IV binds to a pleiotropic protein known as the AT₄ receptor, insulin-regulated aminopeptidase, vasopressinase and oxytocinase (E.C. 3.4.11.3) [5].

A novel angiotensin binding protein, distinct from the known receptors for angiotensin peptides was discovered in the rat brain [6,7]. It is membrane-bound, unmasked in the presence of parachloromercuribenzoic acid (PCMB), not N-glycosylated nor does it behave similarly to most G protein coupled receptors, including the AT₁ receptor, and has an isoelectric point of ~7 [8]. In addition, it has been localized in mouse testis and other structures [9]. These studies led to the identification of this binding site as the metallopeptidase neurolysin (E.C. 3.4.24.16) [10]. Of note, a soluble angiotensin binding protein isolated from porcine liver [11] was later shown to be neurolysin [12], which has 90% homology to rat neurolysin and 91% homology to mouse neurolysin. While the latter study [10] of the angiotensin binding protein was in progress, the present study examined the ability of this protein to interact with other brain peptides, analogs of Ang II and a putative dipeptide inhibitor of neurolysin.

Materials and Methods

Tissue Preparation

Male Sprague-Dawley rats were sacrificed with fluorothane inhalation and brains and testes were harvested, frozen at −20° C and subsequently stored at −80° C. The frozen rat brains and testes were weighed and homogenized for 5–10 seconds in 25 mL of ice-cold, 20 mM NaPO₄, pH 7.4 (hypotonic buffer) by a mechanical homogenizer. Homogenates were centrifuged at 40–48,000× g for 20 minutes and the supernatants were decanted. The membrane pellets were resuspended by homogenization for 5–10 seconds in the assay buffer AM5 (150 mM NaCl, 5mM EDTA, 0.1 mM bacitracin, 50 mM NaPO₄). The homogenates were recentrifuged as before. The final pellet was resuspended in AM5 to give a tissue concentration of 50 mg/mL (initial wet weight). AT₁ and AT₂ blockers, PD 123319 and Losartan (each having a final concentration of 10 µM), and PCMB (final concentration of 100 µM) were added to the tissue homogenate approximately 10 minutes before incubation. Tissue was kept on ice prior to incubation at room temperature. Animal procedures were approved by the IACUCs of Nova Southeastern University and the University of Florida.

Competition Binding Assay Procedure

Competition binding assays were carried out in two steps. For the primary screenings 50 µL membrane preparations were incubated with 0.25 nM ¹²⁵I-SI Ang II and either 10 µM of the test peptide, 3 µM Ang II (all final concentrations) or without any competing ligand “total binding”. Specific binding was determined as “total binding” minus the amount of binding in the presence of 3 µM Ang II “non-specific binding”. All peptides were run in triplicate or more for each assay, and each peptide was assayed in 2 or more separate assays. Peptides were obtained from Phoenix Pharmaceuticals, Bachem, American Peptides, Peninsula, Sigma-Aldrich, and other sources. Ang I Pro¹¹, D-Ala¹² was synthesized by Dr. Gerhard Munskie at Washington State University, HPP SDKP was provided by Dr. Nour-Eddine Rhaleb at Henry Ford Hospital, Detroit MI. In addition, the carboxypeptidase inhibitor DL-mercaptomethyl-3-guanidinoethylthiopropanoic acid (Plummer’s inhibitor) was screened. The tissue homogenate was incubated for one hour at room temperature. Studies carried out in primary cortical neurons obtained from mouse brain indicate that ¹²⁵I-SI Ang II is stable for incubation periods up to 180 min, with > 90% of equilibrium binding occurring by 60 minutes [13]. Membrane bound ¹²⁵I-SI Ang II was collected on glass fiber filters prewetted with 1mg/mL bovine serum albumin using a cell harvester (Model M24R, Brandel, Gaithersburg, MD). Incubation tubes and filters were rinsed three times with 50 mM NaKPO₄, pH 7.4 to minimize non-specific binding to the filters. Bound radioligand maintained on the filter disks was assayed with a COBRA II gamma counter. For three of the peptides that exhibited >50% inhibition of specific ¹²⁵I-SI Ang II binding at 10 µM, as well as proline-isoleucine dipeptide (Pro-Ile), reported to be specific for neurolysin [14], secondary screenings were performed at five different concentrations ranging from 0.1 µM up to 1 mM. Log IC₅₀ was determined by nonlinear regression curve fitting of % inhibition of specific binding data to a one-site competition model algorithm constrained for 100 % specific binding (Prism,Graphpad Software, San Diego, CA). Graphical representation of the data also used Prism software. Values presented are averages ± SEM. Statistical comparisons of Log IC₅₀ values between brain and testis were based on 95% confidence intervals.

Results

Primary screening of thirty-six peptides, endocrine peptides, angiotensin-related peptides, the putative substrates for the AT₄ receptor (vasopressin and oxytocin), the putative neurolysin specific inhibitor, Pro-Ile, and Plummer’s inhibitor, revealed seven peptides with IC₅₀ values < 10 µM for both brain and testis. Three of the peptides were orphanin FQ, Y¹ orphanin FQ, and Y¹⁴ orphanin FQ. Of the three orphanin FQ peptides, only native orphanin FQ was selected for a representative secondary screen. Secondary screenings were also performed on ghrelin, S36057, and Pro-Ile (Figure 1B, C, and D). Primary amino acid sequences for these neuropeptides and angiotensin peptides were aligned for comparison (Table 2). In addition, IC₅₀ values for Ang II and SI Ang II were determined for comparison. Only Ang II and SI Ang II showed IC₅₀ values less than 1 µM (Figure 1A). Percent specific binding of ¹²⁵I-SI Ang II in the presence of each of these peptides was determined at concentrations, ranging from 0.1 µM to 1mM. The IC₅₀ values for the peptides were similar in brain and testis (Table 1) with overlapping 95% confidence intervals for Log IC₅₀ values.

Figure 1. Competition for ¹²⁵I-SI Ang II binding to novel non-AT₁, non-AT₂binding site.

Panel A: IC₅₀ values for Ang II were 246 nM in brain and 356 nM in testis. The IC₅₀ values for SI Ang II were 82.2 nM in brain and 126 nM in testis. The respective Log IC₅₀ values for Ang II in brain and testis were −6.61 ± 0.21 and −6.46 ± 0.17. The respective Log IC₅₀ values for SI Ang II in brain and testis were −7.08 ± 0.18 and −6.90 ± 0.22. Panel B: IC₅₀ values for orphanin FQ were 13.6 µM in brain and 7.95µM in testis. The respective Log IC₅₀ values in brain and testis were −4.87 ± 0.14 and −5.10 ± 0.135. Panel C: IC₅₀ values for ghrelin were 4.15 µM in brain and 2.39 µM in testis. The respective Log IC₅₀ values in brain and testis were −5.38 ± 0.07 and −5.62 ± 0.13. Panel D: IC₅₀ values for Tyr¹ S36057 were 2.07 µM in brain and 1.54 µM in testis. The respective Log IC₅₀ values in brain and testis are −5.68 ± 0.11 and −5.81 ± 0.23. Panel E: Competition curves for Pro-Ile showed no inhibition of ¹²⁵I-SI Ang II binding at concentrations up to 1 mM. Standard error bars for testis are shown in an upward direction and for brain they are shown in a downward direction. Solid symbols represent competition for brain binding, and open symbols represent competition for testis binding.

Table 2.

Primary Amino Acid Sequences of Selected Peptides

Amino Acid Order	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17
Tyr1 S36057	Tyr	*ADO	Arg	Cys	Met	Leu	Gly	Arg	Val	Phe	Arg	Pro	Cys	Trp
Orphanin FQ	Phe	Gly	Gly	Phe	Thr	Gly	Ala	Arg	Lys	Ser	Ala	Arg	Lys	Leu	Ala	Asn	Gln
Tyr1 Orphanin FQ	Tyr	Gly	Gly	Phe	Thr	Gly	Ala	Arg	Lys	Ser	Ala	Arg	Lys	Leu	Ala	Asn	Gln
Tyr14 Orphanin FQ	Phe	Gly	Gly	Phe	Thr	Gly	Ala	Arg	Lys	Ser	Ala	Arg	Lys	Tyr	Ala	Asn	Gln
Dynorphin A (1–8)	Tyr	Gly	Gly	Phe	Leu	Arg	Arg	Ile
Ang I	Asp	Arg	Val	Tyr	Ile	His	Pro	Phe	His	Leu
Ang (1–9)	Asp	Arg	Val	Tyr	Ile	His	Pro	Phe	His
Ang II	Asp	Arg	Val	Tyr	Ile	His	Pro	Phe
Sar1, Ile8 Ang II	Sar	Arg	Val	Tyr	Ile	His	Pro	Ile
Pro-Ile							Pro	Ile
Ghrelin	Gly	Ser	Ser (octanoyl)	Phe	Leu	Ser	Pro	Glu	His	Gln	Lys	Ala	Gln	Gln	Arg	Lys	Glu
Amino Acid Order	18	19	20	21	22	23	24	25	26	27	28
Ghrelin continued	Ser	Lys	Lys	Pro	Pro	Ala	Lys	Leu	Gln	Pro	Arg

^*ADO is 8-amino-3, 6-dioxy octanoyl

Table 1.

Summary of tested neuropeptides and respective IC₅₀ determinations

IC50 Determinations for Thirty-six peptides and related compounds
	Brain (µM)	Testis (µM)
Neuropeptides and Analogs*
Oxytocin	>10	>10
Vasopressin	>10	>10
Orphanin FQ	13.6	7.95
Tyr14 Orphanin FQ	<10	<10
Tyr1Orphanin FQ	<10	<10
Neuropeptide Y (NPY)	>10	>10
Calcitonin Gene-Related Protein (CGRP)	>10	>10
Galanin-like Peptide (GALP)	>10	>10
Met enkephalin	>10	>10
[d-Pen2,d-Pen5] Enkephalin (DPDPE) **	>10	>10
Dynorphin A (1–8)	<10	<10
Pancreatic Polypeptide (PPP)	>10	>10
Somatostatin	>10	>10
Melanin-related
Y1S36057	2.07	1.54
MCH (D Phe13tyr)	>10	>10
Endocrine-related
Ghrelin	4.15	2.39
alpha-Melanocyte Stimulating Hormone (α-MSH)	>10	>10
Cholecystokinin-8 (CCK-8)	>10	>10
Thyroid Releasing Hormone (TRH)	>10	>10
HPP-HOE 140	>10	>10
D-Ala6 des Gly10 LHRH	>10	>10
YFMRF amide	>10	>10
FMRF amide	>10	>10
ArgPheethylamide	>10	>10
ArgPhe	>10	>10
Tyr0 C-type natriuretic peptide (Y0 CNP-22)	>10	>10
Angiotensins
Angiotensin 1 Pro11, D-Ala12	>10	>10
Angiotensin 3–7	>10	>10
Angiotensin 1–9	<10	<10
HPP SDKP	>10	>10
Neurolysin substrates
Orexin B	>10	>10
Urocortin	>10	>10
Neuromedin B	>10	>10
Others
Plummer's inhibitor	>10	>10
Pro-Ile	>10	>10
Tyrosine	>10	>10

Peptides in bold font exhibited > 50% inhibition at 10 µM but were not evaluated in a secondary screen.

^*Functions of many of these peptides can include pituitary and peripheral signaling.

^**Oxidized form

Dynorphin A (1–8) exhibited 69±9% and 82±3% inhibition of ¹²⁵I-SI Ang II binding in brain and testis, respectively. Assuming a one-site competition model the extrapolated IC₅₀ values are 4.5 and 2.2 µM for brain and testis, respectively. Ang (1–9) exhibited 59±3 and 75±2% inhibition of ¹²⁵I-SI Ang II binding in brain and testis, respectively. Assuming a one-site competition model the extrapolated IC₅₀ values are 7.0 and 3.3 µM for brain and testis, respectively.

The 3–7 fragment of Ang II showed less than 10% inhibition of ¹²⁵I-SI Ang II binding in brain and testis at 10 µM. The selective neurolysin inhibitor Pro-Ile also failed to inhibit ¹²⁵I-SI Ang II binding in brain and testis at concentrations up to 1 mM. This value is well above the reported K_I value (90 µM [14]) of Pro-Ile for neurolysin.

Discussion

The main purpose of this study was to further characterize the pharmacological specificity of the novel non-AT₁, non-AT₂ binding site in the rodent brain and testis [6,9] which has been subsequently identified as neurolysin [10]. Competitive binding assays were performed using thirty-six different peptides and a specific peptidase inhibitor to determine whether this binding site is specific for angiotensins or if it has broader ligand/substrate specificity. Primary screening of the thirty-six peptides showed only S36057, orphanin FQ and its analogs, ghrelin, dynorphin A (1–8), and angiotensin (1–9) to inhibit greater than 50% at 10µM in both brain and testis. Secondary screenings performed on three of these peptides indicated IC₅₀ values between 1 and 10 µM, except for orphanin FQ in the brain which was slightly greater than 10 µM.

All tested peptides (Table 1) are grouped according to physiological relevance. Orphanin FQ, also known as nociceptin is a heptadecapeptide which has a primary structure reminiscent of that of opioid peptides[15] and has structural homology with the dynorphin family [16].Orphanin FQ is an endogenous ligand that also inhibits adenylyl cyclase activity. The name nociceptin was derived from its ability to induce enhanced reactivity to noxious thermal stimuli. Studies on rats have shown that orphanin FQ decreases systemic arterial pressure in a dose–related manner [17]. Tyr¹ orphanin FQ, an analog of nociceptin, was also found to behave similarly in terms of vasodepressor activity and potency [16].

Ghrelin is a 28 amino acid peptide that acts as an endogenous orexigenic hormone. This peptide is widely distributed in various tissues, such as the intestines, hypothalamus, and pancreas, along with its mRNA. Studies have also shown ghrelin to have multiple biological functions, including vasodilation and energy homeostasis [18]. S36057 has been classified as a potent ligand for the melanin-concentrating hormone receptor [19]. Melanin-concentrating hormone also has orexigenic actions, as well as reproductive and stress-related signaling activity [20].

The amino acid sequences of orphanin FQ, ghrelin and S36057 do not reveal a recognition motif identical to known neurolysin cleavage sites [21–23], however, these peptides like most neurolysin substrates are primarily basic, and have aromatic amino acids and/or prolines. Given the extensive diversity of amino acids that can mimic the P₃, P₂, P₁, P₁`, P<sub>2</sub>` and P₃` sites of the substrates of neurolysin, these peptides might easily be previously unidentified substrates of neurolysin.

Dynorphin A (1–8) whose first 5 amino acids contain the structure of leucine-enkephalin, is reported to be a substrate of neurolysin [23–26]. Dynorphin A (1–8) is reported to be cleaved at the 5–6 bond by neurolysin to form the endogenous opioid peptide leucine-enkephalin[23,25,26]. As such, only 5 amino acids must be accommodated in the putative substrate binding channel (as described from its crystal structure [22]) distal to the active site.

Angiotensin (1–9) is a likely substrate of neurolysin in view of the high affinity binding of Ang I and Ang II to this protein [6]. Angiotensin (1–9) is formed by the action of angiotensinconverting enzyme 2 (ACE2) albeit at a relatively slow rate compared to the formation of Ang (1–7) from Ang II [27,28]. It is not certain whether neurolysin forms Ang (1–7) or Ang (1–4) and Ang (5–9) from Ang (1–9). If the cleavage site is at the 4–5 bond, then only 4 amino acids must be accommodated in the putative substrate binding channel [22] distal to the active site. In contrast, the 3–7 fragment of Ang II showed negligible inhibition of ¹²⁵I-SI Ang II binding in brain and testis, consistent with a previous observation in mouse brain [7]. This suggests that the arginine² at the P₃ position relative to neurolysin is of critical importance for Ang II binding. It is noteworthy that the acidic carboxy terminal of PCMB has the potential to form an ionic bond with a basic amino acid, such as arginine. This potential for ionic bond formation may be compromised by the aspartic acid¹ of Ang II based upon the higher binding affinities of Ang III and Sar¹, Ile⁸Ang II for this protein relative to Ang II [6,7]. Additionally, the Ang (3–7) and (1–9) data suggests that having a phenylalanine or isoleucine as the carboxy terminal amino acid at the P`₄ position is also critical to high affinity binding of Ang II to neurolysin.

In contrast, several neuropeptides reported to be substrates for neurolysin; urocortin, neuromedin B [29], and somatostatin [23,30] inhibited ¹²⁵I-SI Ang II binding less than 25% at a concentration of 10 µM. Of note, there is one report claiming that human neurolysin does not metabolize somatostatin [31]. Orexin B, which is also reported to be a substrate of neurolysin [29] displayed a slightly greater inhibition of ¹²⁵I-SI Ang II binding, 31 and 41% in brain and testis, respectively.

The putative cleavage sites for urocortin, neuromedin B, somatostatin and orexin B by neurolysin are 13, 9, 6 or 10, and 6 amino acids from the amino terminal of the parent peptide, respectively [23,29]. This portion of these peptides may not be able to fit into the bottom of the ligand binding channel [22] when PCMB is bound to a cysteine sulfhydryl distal to the active site of neurolysin as proposed (Santos et al., submitted for publication).

Pro-Ile has been characterized as a specific albeit low affinity inhibitor of neurolysin. Its K_I is reported to be 90 µM and has been used at concentrations up to 5 mM to selectively inhibit neurolysin activity relative to other endopeptidases [14]. Surprisingly, Pro-Ile did not inhibit ¹²⁵I-SI Ang II binding at concentrations up to 1mM. This observation is reminiscent of previous studies of this non-AT₁, non-AT₂ binding protein in mouse brain in which it was shown that another neurolysin inhibitor, JA-2, also did not compete for ¹²⁵I-SI Ang II binding [7].

Previous studies of the affinity of this PCMB unmasked Ang II binding site for substrates of neurolysin also failed to reveal a specificity characteristic of neurolysin [6]. This suggests that the alteration of neurolysin by PCMB such as to unmask its ability to bind Ang II with high affinity is a pharmacological effect to enhance its affinity for Ang II, while diminishing its affinity for other substrates and Pro-Ile.

Originally described as an inhibitor of angiotensinase C (now known as prolylcarboxypeptidase) [32], PCMB has also been reported to inhibit the catalytic activity of neurolysin [33]. Neurolysin has been shown to metabolize Ang II as well as Ang I [22,23,25,34,35]. The products of Ang II metabolism by neurolysin are the inactive, (1–4) and (5–8) fragments of Ang II [22,23]. The product of Ang I metabolism by neurolysin is Ang (1–7) [25,36]. By degrading Ang II as well as forming Ang (1–7), neurolysin can antagonize the well-known pathophysiological effects of Ang II mediated by the AT₁ receptor [37]. However, there are two reports [26,36] that Ang II is not a substrate of neurolysin despite having K_I values of 4.9 and 2.4 µM for rat and human neurolysin. This suggests that Ang II could be an endogenous inhibitor of neurolysin. Thus, neurolysin may play an important role in the regulation of the RAS. Conversely, neurolysin may be regulated by the RAS, thereby affecting the functionality of other peptidergic systems in the brain and other tissues. Both mechanisms warrant further study of the interactions of the renin-angiotensin system with neurolysin.