Reversible Pegylation Prolongs the Hypotensive Effect of Atrial

Natriuretic peptides (NP), including atrial natriuretic peptide (ANP), induce potent natriuresis and vasodilation and thereby generate hypotension in ...
0 downloads 0 Views 483KB Size
342

Bioconjugate Chem. 2008, 19, 342–348

Reversible Pegylation Prolongs the Hypotensive Effect of Atrial Natriuretic Peptide Maoz Nesher,† Yelena Vachutinsky,‡ Gil Fridkin,§ Yehuda Schwarz,† Keren Sasson,‡ Mati Fridkin,§ Yoram Shechter,‡ and David Lichtstein*,† Department of Physiology, The Hebrew University-Hadassah Medical School, Jerusalem, Israel, and Departments of Biological Chemistry and of Organic Chemistry, The Weizmann Institute of Science, Rehovot, Israel. Received August 5, 2007; Revised Manuscript Received October 30, 2007

Natriuretic peptides (NP), including atrial natriuretic peptide (ANP), induce potent natriuresis and vasodilation and thereby generate hypotension in ViVo. Despite intensive efforts, clinical application of NP as an antihypertensive agent is limited because of their short biological half-life and poor bioavailability. Recently, we have developed a strategy that facilitates slow release of peptides from PEG-peptide inactive conjugates, based on reversible pegylation. Peptides prepared by this approach undergo slow, spontaneous chemical hydrolysis at physiological conditions, releasing the native active peptide/protein drug from the inactive conjugates over prolonged periods. A PEG chain of 30 kDa was linked covalently to the R-amino side chain of the hormone via a MAL-Fmoc-NHS spacer, yielding PEG30-Fmoc-ANP, a prodrug that releases the native hormone upon incubation at physiological conditions. Bolus administration of native ANP to Wistar rats receiving adrenaline yields a short, transitory effect in lowering blood pressure (BP), reaching a maximum at 2 min, and then returning to control values after 12 to 25 min. In contrast, administration of PEG30-Fmoc-ANP lowered BP following a lag period of 50 min, and maintained low BP for a period exceeding 60 min. Saline or PEG30-Fmoc-Alanine were not effective in lowering BP in Wistar rats. These results show that the novel compound, PEG30-Fmoc-ANP, is a reversible pegylated prodrug derivative that facilitates a prolonged BP lowering effect in rats and may be considered as a candidate for development into an antihypertensive drug.

INTRODUCTION Natriuretic peptides (NP), including atrial natriuretic peptide (ANP), brain natriuretic peptide (BNP), and C-type natriuretic peptide (CNP) induce potent natriuresis and vasodilation and thereby generate hypotension in ViVo both in humans and in other mammals (1, 2). ANP stimulates natriuresis by increasing renal glomerular filtration rate and inhibiting Na+ reabsorption at the collecting duct. ANP also suppresses secretion of aldosterone, vasopressin, and renin, thereby stimulating both natriuresis and diuresis (3, 4). In the vasculature, ANP relaxes norepinephrine-, angiotensin II-, and K+-induced vascular smooth muscle contraction causing an immediate hypotensive effect (5). Consequently, ANP plays a central role in controlling blood pressure (BP) and volume regulation (6). Despite intensive efforts, clinical application of the natriuretic peptide family members as antihypertensive agents is limited because of their short biological half-life and poor bioavailability. The t1/2 of ANP is 2 to 5 min in human and even less in the rat (7). Of the three major natriuretic peptide receptors, natriuretic peptide receptors A, B, and C (NPR-A, -B, -C, respectively), NPR-A and NPR-B are guanylyl cyclase-linked, which utilize cGMP as the intracellular messenger (7). The third receptor, NPR-C, is not linked to guanylyl cyclase but appears to act as clearance receptor of the peptides from the circulation (8). Inactivation * Dr. David Lichtstein, Department of Physiology, The Hebrew University - Hadassah Medical School, Jerusalem, 91120, Israel. Phone: 972-2-6758522; Fax: 972-2-6439736; E-mail: [email protected]. † The Hebrew University-Hadassah Medical School. ‡ Department of Biological Chemistry, The Weizmann Institute of Science. § Department of Organic Chemistry, The Weizmann Institute of Science.

of circulating ANP also involves neutral endopeptidase (NEP), a membrane-bound metallopeptidase widely expressed in the vasculature (9). NPR-C and NEP rapidly eliminate ANP from the circulation leading to its short t1/2 and, therefore, limit its suitability as a drug. Nevertheless, intravenous administration of ANP or BNP reduces arterial BP and improves central hemodynamics in patients with chronic heart failure (10, 11). Furthermore, preliminary results indicate that combined inhibition of angiotensin converting enzyme (ACE) and NEP by vasopeptidase inhibitors is an effective strategy in the treatment of hypertension and other cardiovascular diseases including heart failure (12). In addition, Nesiritide (Natrecor; SCIOS, Sunnyvale, CA), a recombinant form of human BNP, has been approved for the treatment of acutely decompensated heart failure by the US Food and Drug Administration since 2001 (13), and ANP is commercially available in Japan under the brand name Carperitide (Astellas Pharma, Fujisawa, Japan) (14). Unfortunately, clinical data suggest that the incidence of angioedema may increase upon vasopeptidase inhibition (15). Moreover, Sackner-Bernstein and co-workers showed that Nesiritide treatment may be associated with increased risk of worsening renal function (16) and even death (17) after treatment of acutely decompensated heart failure. Thus, the development of stableprolonged-acting ANP is of significant therapeutic value. Modifications of peptide/protein can lead to prolonged plasma half-life times. One common method of peptide modification is the covalent attachment of poly(ethylene glycol) (PEG) chain to peptides and proteins (18, 19). Pegylation was shown to improve delivery of drugs and reduce adverse effects (18). Pegylated peptides are shielded from proteolytic enzymes, antibodies, or antigen processing cells (18). The feature gained by pegylation is predominantly a significant lifetime prolongation, in some instance from minutes to hours or even days (20).

10.1021/bc700294w CCC: $40.75  2008 American Chemical Society Published on Web 12/11/2007

Reversible Pegylation of ANP

Bioconjugate Chem., Vol. 19, No. 1, 2008 343

The major obstacle, however, is the inactivation that often takes place upon the covalent attachment of PEG chain to a protein and moreso to peptide drugs. Recently, we developed a strategy, named reversible pegylation, for prolonging the lifetime of peptides and proteins (21). Conjugate prepared by this approach turn into a “prodrug” which undergoes slow, spontaneous hydrolysis at physiological conditions, releasing the native active peptide/protein drug from the inactive conjugate over prolonged periods. This approach was successfully applied to protract the action of exendin-4 (21), interferon R2 (22), peptide YY3–36 (23) and insulin (in preparation). In this study, we initially found that conventional (irreversible) pegylation inactivates ANP. We, therefore, applied the strategy of reversible pegylation to this hormone and investigated its characteristic features in Vitro and its potency of facilitating a long-lasting hypotensive effect in ViVo.

EXPERIMENTAL PROCEDURES Materials. N-R-9-Fluorenylmethoxycarbonyl (Fmoc)-protected amino acids, Fmoc-Tyr(tBu)-preloaded Wang resins, N-hydroxybenzotriazole (HOBt), and benzotriazol-1-yl-oxytripyrrolidinophosphonium hexafluorophosphate (PyBOP) were purchased from Novabiochem (Laufelfingen, Switzerland). N,Ndimethylformamide (DMF), dichloromethane (DCM), N-methyl2-pyrrolidone (NMP), t-butylmethyl ether, and HPLC-grade acetonitrile and trifluoroacetic acid (TFA) were from JT Baker (Phillipsburg, NJ). Dicyclohexylcarbodiimide (DCC) and piperidine were from Merck (Darmstadt, Germany). N-methylmorpholine (NMM) and triethylsilane (TES) were from Fluka (Buchs, Switzerland). PEG30-SH (MPEG-SH-30K) was purchased from Nektar Transforming Therapeutics and contained 0.4 ( 0.03 mol SH per mole solid material as determined by 5,5′-dithiobis(2-nitrobenzoic acid) (DTNB). BSA, SDS, urethane, and dialysis bags were from Sigma (Rehovot, Israel) and adrenaline from Teva (Debrecen, Hungary). All other materials used in this study were of analytical grade. Solid-Phase ANP Synthesis. ANP was synthesized using manual Fmoc-solid-phase methodology on Wang resin preloaded with the Fmoc-protected C-terminal amino acid, i.e., Tyr. DMF and NMP were used as solvents. Double couplings were achieved by reacting 3 Fmoc-AA-OH equivalents with either 3 equiv of PyBOP and 6 equiv of NMM or 3 equiv of HOBT and DCC. The completion of the coupling reaction was indicated by a negative ninhydrin test. NR-Fmoc deprotection was carried out with 20% piperidine in dimethylformamide. Cleavage of the peptides from the solid support and simultaneous side chain deprotection was achieved by a 4 h treatment with trifluoroacetic acid in the presence of triethylsilane, thioanisole, and water (85/ 5/5/5, v/v). After 24 h at room temperature, the cleavage solution was collected and the crude peptides were precipitated from the solution with t-butylmethyl ether at 0 °C. After several washings with ether, the precipitated peptides were dissolved in a solution of water/acetonitrile with 1% TFA and then lyophilized. The peptide containing cysteine residues was cyclized by intramolecular oxidation. This peptide (0.2 mg/mL) was dissolved at high dilution in 0.1 M ammonium acetate containing potassium ferricyanide (0.01 mM; 1.4 equiv) and stirred overnight. Reduction in the content of cysteine moieties was followed using DTNB. The crude peptide was purified by reversed-phase HPLC (see below) on a semipreparative C18 column and characterized by ESMS and amino acid analysis. The product appeared as a single peak on analytical HPLC chromatography (Rt ) 3.3 min), having identical Rt value as that of commercially available human ANP, indicating that the synthesis, cyclization, and purification yielded the desired product. Amino acid analysis following acid hydrolysis showed

Figure 1. Structure of 9-hydroxymethyl-2-(amino-3-maleimidopropionate)-7-N-hydroxysuccinimide (MAL-Fmoc-NHS) (A). Schematic presentation of the steps involved in the preparation of PEG30-FmocANP (B).

the expected amino acid composition. Mass spectrometry analysis (see below) revealed a mass [M - H]+ of 3081.21 ( 0.43 Da. Calcd. 3081.44 Da. ANP Conjugates Preparation.Preparation of PEG30-FmocANP. MAL-Fmoc-NHS (Figure 1A) is a heterobifunctional agent consisting of fluorenemethoxycarbonyl N-hydroxysuccinimide ester that reacts with amino groups of peptides and proteins. A maleimide group, attached to the fluorenyl backbone, enables its coupling to any sulfhydryl-containing compound. Figure 1B shows the optimal procedure found for preparing PEG30-Fmoc-ANP. Accordingly, to a stirred solution of ANP (1 mg) in 100 µL of 0.1 M phosphate buffer, pH 7.2 (3.24 mM), 10 mg of PEG30-SH in 100 µL phosphate buffer was added to obtain a final concentration of 3.33 mM. MAL-Fmoc-NHS (15 µL) from a fresh solution of MAL-Fmoc-NHS (10 mg/mL in DMF, 150 µg) was then added. The reaction was carried out for 4 h followed by dialysis for 2 days against distilled H2O to remove residual DMF, phosphate buffer, and free ANP. Alternatively, the product was purified by HPLC. The product was characterized by MALDI-TOF mass spectrometry. Effective molecular weight of the conjugate was obtained by SDS gel electrophoresis (see below). The amount of ANP in the conjugate was determined by acid hydrolysis of a 20 µL of 2 mg/mL aliquot, followed by amino acid analysis. ANP concentration was calculated according to alanine (1 residue), tyrosine (1 residue), isoleucine (1 residue), leucine (2 residues), and phenylalanine (2 residues). MALDI-TOF mass spectral analysis showed 1:1 PEG30-Fmoc/ANP stoichiometry. The molecular mass ([M - H]+) of synthesized ANP was 3081.21 Da as determined by ESMS. PEG30-SH had an average molecular mass of 32 300 Da. The spacer molecule following conjugation had a MW of 387 Da, and the conjugate yielded a major peak corresponding to a molecular mass of 35 675 Da. PEG30-Fmoc-ANP migrates on analytical HPLC column (see below) as a wide peak with Rt value of 43 min. NonreVersible Pegylation of ANP. A stable, nonreversible PEG-ANP conjugate was prepared using the same procedure described for PEG30-Fmoc-ANP, except for the use of 3-maleimidobenzoate-N-hydroxysuccinimide ester (MIB-NHS, Pierce, Rockford, USA) to link PEG30-SH to ANP.

344 Bioconjugate Chem., Vol. 19, No. 1, 2008

High Performance Liquid Chromatography (HPLC) and Mass Spectrometry. Reverse-phase HPLC was performed with a Spectra-Physics SP8800 liquid chromatography system (Spectro-Physics, San Jose, CA) equipped with an Applied Biosystem 757 variable-wavelength absorbance detector. The column effluents were monitored by UV absorbance at 220 nm, and chromatograms were recorded on a chromJet integrator (Thermo-Separation, Riviera Beach, FL). Compounds were analyzed by analytical reversed-phase HPLC using an RP-18, 100 × 4.6 mm Chromolith column (Merck, Darmstadt, Germany). HPLC prepacked columns were either Vydac RP-18, 250 × 22 mm, bead size 10 µm for ANP purification, or Vydac RP-4, 250 × 22 mm, bead size 10 µm, for pegylated ANP derivative purification (Bucher Biotec AG, Basel, Switzerland). Linear gradients were formed between solution A (0.1% TFA in H2O) and solution B (0.1% TFA in acetonitrile-H2O 75:25). The analytical column was eluted with a binary gradient of 10–100% solution B over 14 min at a flow rate of 3 mL. Purifications were achieved with a 10–100% B gradient over 50 min, with a flow of 10.0 mL/min. Mass spectra were determined using matrix-assisted laserdesorption/ionization time-of-flight (MALDI-TOF) mass spectroscopy (Micromass UK Ltd.) and electrospray ionization mass spectra (ESMS) techniques, performed on a Micromass Platform LCZ 4000 (Manchester, UK). Amino acid composition was analyzed following acid hydrolysis in 6 N HCl at 110 °C for 24 h by Dionex amino acid analyzer (Dionex Corporation, CA. USA). SDS-PAGE Electrophoresis. To determine effective molecular weight of PEG30-Fmoc-ANP/PEG-ANP conjugates, SDS-PAGE was performed using 15% polyacrylamide slab gel according to the Laemmeli method. 10 µg of each sample and a standard protein solution were boiled for 3 min, centrifuged, and loaded on a SDS-PAGE. Pegylated ANPs were specifically stained with a barium iodide solution according to the method of Kurfurst (24). This specific staining is based on the formation of a barium iodide complex with the poly(ethylene glycol) molecule. After electrophoresis, the gel was soaked in a 5% glutaraldehyde solution for 15 min at room temperature for fixation and then immersed in 20 mL of 0.1 M perchloric acid for 15 min. After these washes, 10 mL of a 5% barium chloride solution was added, followed by 4 mL of 0.1 M iodine solution. The stained PEG bands appeared within a few minutes. After 15 min, the staining solution was replaced and the gel was put into water. Biological Methods.[125I]-ANP Binding to HeLa Cells. Cell Culture. NPR-A expressing HeLa cells were obtained from Professor Haim Garty (Departments of Biological Chemistry, The Weizmann Institute of Science, Rehovot, Israel) and grown at 35 °C in a 5% CO2 humidified incubator in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal calf serum, penicillin, and streptomycin (100 U/mL of each). The binding assay was performed on subconfluent cell monolayers in 24-well dishes at a density of 4 · 105 cells/well preincubated for 24 h. [125I]-ANP Binding. ANP was iodinated by the chloramine T procedure of Hunter and Greenwood (25). HeLa cell monolayers were washed twice with PBS buffer (pH 7.4) containing 1% bovine serum albumin and then incubated in the same buffer containing [125I]-ANP (100 000 cpm/well) and increasing concentrations (0.74 to 37 nM) of either native ANP or PEGylated ANP derivatives for 2 h at 4 °C. Nonspecific binding was measured in the presence of excess, unlabeled ANP (1.5 µM). The cells were then washed twice with ice-cold buffer and solubilized with 0.2% w/v sodium dodecylsulphate (SDS). Following incubation (37 °C for 30 min), the solubilized cells were transferred to counting vials, and radioactivity was counted

Nesher et al. Table 1. Molecular Weight of PEG30-Fmoc-ANP, as Determined by MALDI-TOF and by SDS Gel Electrophoresisa mass (g/mol) (calculated)

[M + H+] (g/mol) (found by MALDI-TOF)

effective MW according to SDS gel electrophoresis

35 768

35 675

∼68 000

a

Molecular mass was determined by MALDI-TOF and by SDS gel electrophoresis as described in the Experimental Procedures section.

(γ-counter, Packard Cobra, GMI, USA). All experiments were performed in duplicate, and each experiment was repeated at least three times. Specific binding is defined as total radioactive content (in the absence of native ANP) minus that obtained at high (1.5 µM) concentration of unlabeled ANP. Receptor binding capacity was determined by calculating the concentration of unlabeled ANP or its derivative that was required to displace 50% of specifically bound [125I]-ANP. In Vivo Studies.Animals. Experiments were performed on male Wistar rats weighing 300–350 g. The rats were housed according to a 12 h light/dark cycle and were allowed at least a 3 day acclimatization period with normal rat chow and tap water ad libitum. All procedures were carried out according to the guidelines of the Hebrew University-Hadassah Medical School Animal Care Committee for the use and care of laboratory animals. Experimental Methods. Animals were anesthetized with urethane (1 g/kg, i.p.) and placed on a heated pad. Rectal temperature was adjusted to 37.5–38 °C by a rectal probe. Following tracheotomy (PE260 catheter), the two femoral veins and the right femoral artery were cannulated with a PE50 heparanized tube. Blood pressure (BP) was recorded from the right femoral artery using SP 844 pressure transducer (MEMSCAP, Skoppum, Norway) coupled to an amplifier (Laboratory linc V, Coulbourn instruments, Allentown, PA). The analog signal was converted to a digital recording by a PowerLab 16SP integration system, and the data were stored on a personal computer by Chart 4.0 software (Adinstruments, Castle Hill, Australia). Heart rate (HR) was calculated from the BP beats. All animals received an infusion of 3.3 µg/kg/min of adrenaline throughout the experiment using syringe pump (Stoelting, IL). After 30 min, the test compounds (3 µg/kg ANP, 100 µg/kg ANP, 200 µg/kg PEG30-Fmoc-alanine, and PEG30-Fmoc-ANP 96 ( 7 µg covalently linked ANP/kg) were administered by a bolus injection through the second femoral vein cannula in 100 µL saline. In some experiments, 0.83 µg/kg ANP, in 30 µL saline/min, was administered by infusion. The animals were monitored for 120 min after the treatments and then sacrificed. Animals manifesting unstable BP, irregular heart beats, or breathing abnormalities during the experiments were excluded. Statistical Analysis. Mean arterial blood pressure (MAP), systolic blood pressure (SBP), and HR were calculated by Chart 4.0 software. Statistical analysis was performed for the changes in BP parameters between sequential measurements and time 0 (just before compound administration) of the experiment in 2 min intervals. This allowed for direct comparison of the responses to treatment between groups when baselines differed (26). One-way analysis of variance (ANOVA) was used to compare differences among groups, and two-way ANOVA for repeated measures, followed by Tukey posthoc test. Student’s paired t test was used to compare differences in the same group. SPSS 11.5 for windows (SPSS Inc., Chicago, IL) was used to perform the statistical analyses.

RESULTS ANP Conjugates Syntheses. Table 1 summarizes several characteristic features of the PEG30-Fmoc-ANP conjugate. ANP contains N-terminal R-amino group available for acylation by

Reversible Pegylation of ANP

Bioconjugate Chem., Vol. 19, No. 1, 2008 345

Table 2. Receptor Binding Capacities of ANP and Its Conjugatesa ANP PEG30-ANP (nonreversibly PEGylated) PEG30-Fmoc-ANP PEG30-Fmoc-ANP incubated in PBS-1% BSA pH ) 7.4 (2 d, 37 °C) PEG30-Fmoc-ANP incubated in 0.1 M borate buffer, pH ) 8.5 (2 d, 37 °C)

IC50 (nM)

biological potency (%)

1.31 ( 0.01 127b ( 7.4

100 1.03

89b ( 10.37 12b ( 0.51

1.47 10.91

11b ( 0.44

11.9

a 125

[ I]-ANP binding to HeLa cells was performed as described in Experimental Section. b ANP concentration was calculated based on the amount of covalently linked ANP in the conjugate, as measured by amino acid analysis following acid hydrolysis. IC50 are expressed as mean ( SD of 4 experiments. Biological potency was calculated as (IC50 of ANP/IC50 of test compound) × 100.

MAL-Fmoc-NHS. Since our heterobifunctional agent (MALFmoc-NHS) is highly specific to amino side chains in non-SHcontaining peptides and proteins (21–23) and ANP is a nonlysine-containing peptide agonist (3), only the N-terminal R-amino group of ANP is available for acylation. The synthetic procedure applied (Experimental Procedures section) was followed by extensive dialysis and HPLC purification to remove any residual ANP expected, thus yielding a monomodified homogeneous conjugate. Indeed, MALDI-TOF mass spectral analysis shows a main peak having 1:1 PEG30-Fmoc/ANP stoichiometry. The molecular mass of synthesized ANP was 3081 Da as determined by ESMS. PEG30-SH had an average molecular mass of 32 300 Da. The spacer molecule following conjugation has a MW of 387 Da, and the conjugate yielded a major peak corresponding to a molecular mass of 35 675 Da. Acid hydrolysis and amino acid analysis of PEG30-Fmoc-ANP revealed the content of 31 ( 3 µg ANP rather than 86 µg ANP per mg solid material as has been calculated for this molecular weight. This suggests that our preparation also contains PEG30Fmoc that has not been linked to ANP. Further dosage calculations, however, were based on the value of 31 ( 3 µg ANP per mg solid material, namely, that determined by amino acid analysis following acid hydrolysis. Receptor Binding Capacity of ANP Conjugates. HeLa cells bind [125I]-ANP with high affinity and the displacement of the ligand by unlabeled ANP yielded an IC50 value of 1.3 ( 0.096 nM (data not shown). Previous studies demonstrated that this binding is associated with elevation in the cellular cGMP level (27, 28). Table 2 summarizes the receptor binding capacities of the native and pegylated derivatives of ANP in HeLa cells. Relative IC50 values obtained revealed that both conventional pegylated ANP and PEG30-Fmoc-ANP have ∼1% of the receptor binding affinity of the native peptide. PEG30Fmoc-ANP, however, undergoes reactivation upon incubation at physiological conditions. This was valid for PEG30-FmocANP incubated either in 0.1 M borate buffer (pH 8.5, 37 °C) or in PBS buffer containing 1% w/v BSA (pH 7.4, 37 °C). Release of ANP upon Incubation at Physiological Conditions. PEG30-Fmoc-ANP (322 µg/mL, 10 µg covalently linked ANP/mL) was incubated in 0.1 M borate buffer, pH 8.5, at 37 °C (see Discussion). Aliquots were withdrawn at different time points and analyzed both for the amount of ANP released and for receptor-binding potency using the [125I]-ANP displacement assay in HeLa cells (Figure 2). Under these incubation conditions, peptides are hydrolyzed from PEG-Fmoc-peptide conjugates at a rate that is nearly identical to that obtained in normal human serum at 37 °C (21, 22). Upon incubation, ANP has been hydrolyzed from the conjugate in a nearly linear fashion with the concomitant increase in receptor binding potency (Figure 2). About 7% (0.78 ( 0.03 µg) of ANP has been

Figure 2. Rate of ANP release from PEG30-Fmoc-ANP. PEG30-FmocANP (10 µg covalently linked ANP) was incubated in borate buffer (pH 8.5) at 37 °C. Aliquots were withdrawn at the indicated time points and analyzed for the amount of ANP released (O) and for receptor binding capacity (•) (Experimental Procedures section). Binding capacity is calculated as the IC50 for the ANP released, versus the IC50 of the same quantity of the native hormone. The data are presented as mean ( SD of 4 experiments.

released on day 1 (Figure 2). Based on this value and preliminary results, we calculated that 1 mg of PEG30-FmocANP (namely, 31 ( 3 µg covalently linked ANP) per rat or more is required for the in ViVo experiments. Effect of ANP on Mean Arterial and Systolic BP. Following infusion of adrenaline (3.3 µg/kg/min) for 30 min, rats were injected with native ANP at a dose of 3 and 100 µg/kg. These treatments elicited significant reductions in mean arterial and systolic BP (Figure 3A,B). As expected, ANP effects were transient, and even at the high dose (100 µg/kg), BP returned to control values 20–25 min after the injection. These reductions in mean arterial and systolic BP were accompanied with insignificant, minor changes in diastolic BP and HR (data not shown). These results are in agreement with many studies demonstrating the hypotensive transient effects of ANP (29, 30). Notably, a fall in BP at the lower ANP dose (3 µg/kg) was not detected in rats that were not pretreated with adrenaline (data not shown). Therefore, the increase in sensitivity to ANP and its derivatives following adrenaline pretreatment has been subsequently applied. Prolonged Hypotensive Effects Following Administration of PEG30-Fmoc-ANP. Mean arterial and systolic BP of control animals receiving only adrenaline infusion (3.3 µg/kg/ min) decreased slightly by about 10 mmHg within 2 h (Figure 4A,B). This reduction is probably due to time-dependent desensitization of the adrenergic receptors as reported previously (31). The administration of PEG30-Fmoc-ANP at a dose of 96 ( 7 µg covalently linked ANP/kg body weight did not change the mean arterial and systolic BP until about 50 min after injection. Mean arterial and systolic BPs decreased markedly and remained low over a period exceeding 60 min (Figure 4A,B). For comparison, the effect of continous infusion of native ANP at a rate of 0.83 µg/kg/min is demonstrated as well (Figure 4A,B). As shown previously (30, 32), this treatment yields an immediate effect on lowering both mean and systolic BP in rats, which persists as long as ANP is being infused. Indeed, as reported here, a marked reduction in BP is obtained over a period of 10 to 95 min. Bolus injection of PEG30-Fmoc-alanine at a dose of 200 µg/kg body weight showed insignificant effect on BP that did not differ from that obtained in a saline-treated group (data not shown). Thus, a single administration of PEG30Fmoc-ANP at a dose of ∼100 µg covalently linked ANP/kg facilitates a marked and prolonged BP lowering action in rats,

346 Bioconjugate Chem., Vol. 19, No. 1, 2008

Nesher et al.

Figure 3. Effect of ANP on BP parameters in anesthetized rats. Animals were anesthetized with urethane, and BP was recorded from the right femoral artery as described in the Experimental Procedures section. All animals received an infusion of 3.3 µg/kg/min of adrenaline throughout the experiment. After 30 min (indicated by an arrow), saline (O), 100 µg/kg ANP (•), and 3 µg/kg ANP (2) were administered intravenously by 100 µL bolus injection. Values are expressed as mean ( SE (n ) 6). *P < 0.05 vs control; **P < 0.01 vs control; †P < 0.05 vs baseline; ††P < 0.01 vs baseline.

Figure 4. Effect of PEG30-Fmoc-ANP on BP in anesthetized rats. Animals were anesthetized with urethane and BP was recorded from the right femoral artery as described in the Experimental Procedures section. All animals received an infusion of 3.3 µg/kg/min of adrenaline throughout the experiment. After 30 min (indicated by an arrow), the animals received 100 µL bolus injection of saline (O), 96 ( 7 µg/kg covalently linked PEG30-Fmoc-ANP (2), or infusion of 0.83 µg/kg/min ANP (•). Values are expressed as mean ( SE (n ) 6). *P < 0.05 vs control; **P < 0.01 vs control; †P < 0.05 vs PEG30-Fmoc-ANP; ††P < 0.01 vs PEG30-Fmoc-ANP.

which starts at 50 min after administration and resembles the pattern obtained by continuous infusion of ANP.

DISCUSSION Since the pioneering work of de Bold and his colleagues (29), demonstrating the natriuretic response to atrial extracts in rats, many attempts were made to utilize the active substance from that extract (i.e., ANP) as a natural hypotensive agent. It has become clear that continuous infusion of ANP yields natriuresis, diuresis, and hypotension in animal models and in humans (1, 7), indicating that this natural peptide may have a significant therapeutic value. However, ANP is a short-lived species in ViVo, as it undergoes proteolysis in the circulatory system by NEP and NPR-C and is rapidly cleared by the kidneys (3, 8, 9). Therefore, the hypotensive effect of ANP is a transient phenomenon (29). The rapid degradation of ANP in the serum poses a serious limitation on its practical use. Therefore, developing a long-acting version of this hormone is a prerequisite for its therapeutic application in the future. The covalent attachment of PEG to peptides/proteins (pegylation) was shown to dramatically prolong the lifetime of the PEG-peptide conjugate and shield them from proteolysis in ViVo (18–20, 33). Therefore, we first attempted this strategy by covalently linking PEG30-SH to ANP through the nonreversible linker MIB-NHS. A PEG chain of 30 kDa has been selected,

since lower molecular weight PEG chains are considerably less effective in preventing clearance by kidney filtration (34). Indeed, PEG30-Fmoc and PEG30-Benz-ANP, having calculated molecular weights of ∼36 kDa, migrated as a ∼68 kDa species on SDS gel electrophoresis (Table 1 A), indicating a substantial enlargement in the hydrodynamic volume and therefore in the effective molecular weight of these conjugates. This value is very close to the ∼70 kDa cutoff for glomerular filtration of native globular proteins (35). In spite of those profound advantages, a major obstacle of conventional pegylation is the inactivation that often takes place, particularly with regard to short peptide agonists upon derivatization. A proportionally larger section of low MW peptide agonists participate in and must be accessible to receptor binding. Indeed, as found here, irreversible pegylation of ANP nearly fully abolished its receptor binding to HeLa cells (Table 2). The remaining alternative was therefore to apply the strategy of reversible pegylation. In this study, we have linked PEG30SH to the single R-amino side chain of ANP via the heterobifunctional spacer MAL-Fmoc-NHS (Figure 1A). The latter was designed on the basis of the Fmoc principle by which the peptide is released from the conjugate in its native active form. The Fmoc group is used in organic and peptide synthesis for the protection of amino groups and can be removed following basic

Reversible Pegylation of ANP

treatment, such as with piperidine, within several minutes (36, 37). In neutral, aqueous solutions, Fmoc moieties undergo slow, spontaneous hydrolysis, resulting in the regeneration of the native proteins. Since ANP is a non-lysine-containing peptide, an homogeneous 1:1 monomodified PEG-Fmoc-ANP conjugate, where the PEG chain is linked exclusively to the R-amino function of the hormone, has been obtained (Table 1). As expected, PEG30-Fmoc-ANP is an inactive conjugate as well, and therefore a classical prodrug. This derivative, however, releases ANP and is reactivated upon incubation at physiological conditions (Figure 2). Although this process takes place at a slow rate, receptor binding capacity showed a nearly 7-fold increase within two days of incubation at pH 8.5 and 37 °C (Table 2). Previously, we have shown that the rate of hydrolysis of Fmoc-derived species at pH 8.5, 37 °C, is nearly similar to that obtained at pH 7.4 in the presence of physiological conditions of albumin. We did attribute this finding to the catalytic function of albumin (21–23). A linear relationship was obtained between receptor binding potency and the amount of ANP detached from the conjugate (Figure 2). Following the completion of these in Vitro studies, we investigated whether this conjugate could facilitate a prolonged effect in ViVo. Administration of ANP at a dose of 0.1–3 µg/kg to normotensive rats failed to lower BP at any time point following injection (data not shown). Thus, native rats do not provide a suitable experimental pharmacological model to study ANPinduced hypotension. We therefore took advantage of the observation that ANP effects can be intensified by catecholamines (38–40). Accordingly, we infused adrenaline to all our animal groups to elevate BP and sensitize the effect of ANP. Indeed, the administration of 3 µg/kg of ANP, which by itself does not affect BP in naive animals, reduced it significantly in rats receiving constant infusion of adrenaline (Figure 3). Notably, this in ViVo experimental system appears to reflect well the short half-life of ANP (t1/2 ) 2–5 min, (7)). Even with the intravenous administration of a relatively high dosage of ANP (100 µg/kg, 30 µg/rat), BP returned to control levels within 20–25 min following administration (Figure 3). This is in complete agreement with other studies (29, 30). The protracted action of PEG30-Fmoc-ANP in lowering BP is demonstrated in Figure 4. Note should be taken that our in ViVo experimental system limits the measurement of BP up to a period of 2 h. BP, however, has reached its lowest value 2 h after PEG30-Fmoc-ANP administration. Extrapolation of these data reveals that a low level of BP could be maintained for a period well exceeding ∼10 h as we previously documented with related PEG30-Fmoc-peptide analogues (21, 23). Rough calculation based on (1) the administration of 30 µg covalently linked ANP/rat, with a 10× dilution shortly after administration, and (2) the rate of ANP released found in Vitro (Figure 2) reveals that a level of about 6–10 ng/mL of free ANP is expected to be maintained in the circulatory system at any time point following a lag period of ∼1 h. This ANP level is well within the Kd value for in Vitro ANP binding affinity to intact cellular systems, (Kd ≈ 2 nM, (28)).

CONCLUSION In summary, we managed to overcome several obstacles that prevent the designing of prolonged, continuous, therapeutically active ANP. The strategy of conventional pegylation has been replaced by reversible pegylation, an approach that resembles a continuous ANP infusion by the continuous release of the hormone over a prolonged period following administration. Our results show that the novel compound PEG30-Fmoc-ANP release ANP under physiological conditions prolonging the action of the active peptide.

Bioconjugate Chem., Vol. 19, No. 1, 2008 347

ACKNOWLEDGMENT We thank Dr. U. Shpolansky for his help in setting up the in ViVo experimental system. M.F. is the Lester Pearson Professor of Protein Chemistry. Y.S. is the incumbent of the C.H. Hollenberg Chair in Metabolic and Diabetes Research, established by the friends and associates of Dr. C.H. Hollenberg of Toronto, Canada.

LITERATURE CITED (1) Silver, M. A. (2006) The natriuretic peptide system: kidney and cardiovascular effects. Curr. Opin. Nephrol. Hypertens. 15, 14–21. (2) Nakayama, T. (2005) The genetic contribution of the natriuretic peptide system to cardiovascular diseases. Endocr. J. 52, 11– 21. (3) Levin, E. R., Gardner, D. G., and Samson, W. K. (1998) Natriuretic peptides. N. Engl. J. Med. 339, 321–328. (4) McGrath, M. F., de Bold, M. L., and de Bold, A. J. (2005) The endocrine function of the heart. Trends Endocrinol. Metab. 16, 469–477. (5) Ahluwalia, A., MacAllister, R. J., and Hobbs, A. J. (2004) Vascular actions of natriuretic peptides. Cyclic GMP-dependent and -independent mechanisms. Basic Res. Cardiol. 99, 83–89. (6) Antunes-Rodrigues, J., de Castro, M., Elias, L. L., Valenca, M. M., and McCann, S. M. (2004) Neuroendocrine control of body fluid metabolism. Physiol. ReV. 84, 169–208. (7) Potter, L. R., Abbey-Hosch, S., and Dickey, D. M. (2006) Natriuretic peptides, their receptors, and cyclic guanosine monophosphate-dependent signaling functions. Endocr. ReV. 27, 47– 72. (8) Anand-Srivastava, M. B. (2005) Natriuretic peptide receptor-C signaling and regulation. Peptides 26, 1044–1059. (9) Corti, R., Burnett, J. C., Jr., Rouleau, J. L., RuschitzkaF., and Luscher, T. F. (2001) Vasopeptidase inhibitors: a new therapeutic concept in cardiovascular disease. Circulation 104, 1856–1862. (10) Maisel, A. S. (2003) Nesiritide: a new therapy for the treatment of heart failure. CardioVasc. Toxicol. 3, 37–42. (11) Suwa, M., Seino, Y., Nomachi, Y., Matsuki, S., and Funahashi, K. (2005) Multicenter prospective investigation on efficacy and safety of carperitide for acute heart failure in the ‘real world’ of therapy. Circ. J. 69, 283–290. (12) Quaschning, T. (2005) Vasopeptidase inhibition for blood pressure control: emerging experience. Curr. Pharm. Des. 11, 3293–3299. (13) Burger, A. J., and Burger, M. R. (2002) Nesiritide (Scios). IDrugs 5, 703–709. (14) Vesely, D. L. (2006) Which of the cardiac natriuretic peptides is most effective for the treatment of congestive heart failure, renal failure and cancer? Clin. Exp. Pharmacol. Physiol. 33, 169– 176. (15) Messerli, F. H., and Nussberger, J. (2000) Vasopeptidase inhibition and angio-oedema. Lancet 356, 608–609. (16) Sackner-Bernstein, J. D., Skopicki, H. A., and Aaronson, K. D. (2005) Risk of worsening renal function with nesiritide in patients with acutely decompensated heart failure. Circulation 111, 1487– 1491. (17) Sackner-Bernstein, J. D., Kowalski, M., Fox, M., and Aaronson, K. (2005) Short-term risk of death after treatment with nesiritide for decompensated heart failure: a pooled analysis of randomized controlled trials. JAMA 293, 1900–1905. (18) Veronese, F. M., and Pasut, G. (2005) PEGylation, successful approach to drug delivery. Drug DiscoVery Today 10, 1451– 1458. (19) Werle, M., and Bernkop-Schnurch, A. (2006) Strategies to improve plasma half life time of peptide and protein drugs. Amino Acids 30, 351–367. (20) Roberts, M. J., Bentley, M. D., and Harris, J. M. (2002) Chemistry for peptide and protein PEGylation. AdV. Drug DeliVery ReV. 54, 459–476.

348 Bioconjugate Chem., Vol. 19, No. 1, 2008 (21) Tsubery, H., Mironchik, M., Fridkin, M., and Shechter, Y. (2004) Prolonging the action of protein and peptide drugs by a novel approach of reversible polyethylene glycol modification. J. Biol. Chem. 279, 38118–38124. (22) Peleg-Shulman, T., Tsubery, H., Mironchik, M., Fridkin, M., Schreiber, G., and Shechter, Y. (2004) Reversible PEGylation: a novel technology to release native interferon alpha2 over a prolonged time period. J. Med. Chem. 47, 4897–4904. (23) Shechter, Y., Tsubery, H., Mironchik, M., Rubinstein, M., and Fridkin, M. (2005) Reversible PEGylation of peptide YY3–36 prolongs its inhibition of food intake in mice. FEBS Lett. 579, 2439–2444. (24) Kurfurst, M. M. (1992) Detection and molecular weight determination of polyethylene glycol-modified hirudin by staining after sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Anal. Biochem. 200, 244–248. (25) Hunter, W. M., and Greenwood, F. C. (1962) Preparation of iodine-131 labelled human growth hormone of high specific activity. Nature (London) 194, 495–496. (26) Rybczynska, A., Boblewski, K., Lehmann, A., Orlewska, C., Foks, H., Drewnowska, K., and Hoppe, A. (2005) Calcimimetic NPS R-568 induces hypotensive effect in spontaneously hypertensive rats. Am. J. Hypertension 18, 364–371. (27) Kort, J. J., and Koch, G. (1990) Receptors for atrial natriuretic peptide (ANP) and cyclic GMP responses in HeLa cells. Biochem. Biophys. Res. Commun. 168, 148–154. (28) Watt, V. M., and Yip, C. C. (1989) HeLa cells contain the atrial natriuretic peptide receptor with guanylate cyclase activity. Biochem. Biophys. Res. Commun. 164, 671–677. (29) de Bold, A. J., Borenstein, H. B., Veress, A. T., and Sonnenberg, H. (1981) A rapid and potent natriuretic response to intravenous injection of atrial myocardial extract in rats. Life Sci. 28, 89–94. (30) Sasaki, A., Kida, O., Kangawa, K., Matsuo, H., and Tanaka, K. (1985) Hemodynamic effects of alpha-human atrial natriuretic polypeptide (alpha-hANP) in rats. Eur. J. Pharmacol. 109, 405– 407.

Nesher et al. (31) Guimaraes, S., and Moura, D. (2001) Vascular adrenoceptors: an update. Pharmacol. ReV. 53, 319–356. (32) Hirata, Y., Ishii, M., Sugimoto, T., Matsuoka, H., Fukui, K., Sugimoto, T., Yamakado, M., Tagawa, H., Miyata, A., and Kangawa, K. (1988) Hormonal and renal effects of atrial natriuretic peptide in patients with secondary hypertension. Circulation 78, 1401–1410. (33) Shepherd, J., Jones, J., Takeda, A., Davidson, P., and Price, A. (2006) Adefovir dipivoxil and pegylated interferon alfa-2a for the treatment of chronic hepatitis B: a systematic review and economic evaluation. Health Technol. Assess. 10, 1–183, iii-iv, xi-xiv. (34) Rosendahl, M. S., Doherty, D. H., Smith, D. J., Carlson, S. J., Chlipala, E. A., and Cox, G. N. (2005) A long-acting, highly potent interferon alpha-2 conjugate created using site-specific PEGylation. Bioconjugate Chem. 16, 200–207. (35) Caliceti, P., and Veronese, F. M. (2003) Pharmacokinetic and biodistribution properties of poly(ethylene glycol)-protein conjugates. AdV. Drug DeliVery ReV. 55, 1261–1277. (36) Carpino, L., and Han, G. (1972) The 9-fluorenylmethoxycarbonyl amino-protecting group. J. Org. Chem. 37, 3404–3409. (37) Carpino, L., and Han, G. (1970) The 9-fluorenylmethoxycarbonyl function, a new base-sensitive amino-protecting group. J. Am. Chem. Soc. 92, 5748–5749. (38) Yasujima, M., Abe, K., Kohzuki, M., Tanno, M., Kasai, Y., Sato, M., Omata, K., Kudo, K., Tsunoda, K., and Takeuchi, K., et al. (1985) Atrial natriuretic factor inhibits the hypertension induced by chronic infusion of norepinephrine in conscious rats. Circ. Res. 57, 470–474. (39) Badalamenti, S., Borroni, G., Lorenzano, E., Incerti, P., and Salerno, F. (1992) Renal effects in cirrhotic patients with avid sodium retention of atrial natriuretic factor injection during norepinephrine infusion. Hepatology 15, 824–829. (40) Parkes, D. G., Coghlan, J. P., McDougall, J. G., and Scoggins, B. A. (1990) Effects of atrial natriuretic factor on pressor responsiveness to angiotensin II, norepinephrine, and vasopressin in conscious sheep. J. CardioVasc. Pharmacol. 15, 16–21. BC700294W