Atrial Natriuretic Peptide-Fc, ANP-Fc, Fusion Proteins - ACS Publications

Jan 22, 2012 - ... peptides and therapeutic potential in the metabolic syndrome ... Yan, Yao, Xu, Martin, Zhang, Chu, dela Cruz-Chuh, O'Donohue, Li, D...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/bc

Atrial Natriuretic Peptide-Fc, ANP-Fc, Fusion Proteins: Semisynthesis, In Vitro Activity and Pharmacokinetics in Rats Adam R. Mezo,*,† Kevin A. McDonnell,‡ Susan C. Low,‡ Jeff Song,‡ Tom J. Reidy,† Qi Lu,† John V. Amari,† Todd Hoehn,† Robert T. Peters,† Jennifer Dumont,† and Alan J. Bitonti† †

Biogen Idec Hemophilia, 9 Fourth Avenue, Waltham, Massachusetts 02451, United States S Supporting Information *

ABSTRACT: Atrial natriuretic peptide (ANP) may be a useful molecule for the treatment of cardiovascular diseases due to its potent natriuretic effects. In an effort to prolong the short in vivo half-life of ANP, fusions of the peptide to the Fc domain of IgG were generated using a semisynthetic methodology. Synthetic ANP peptides were synthesized with thioesters at either the N- or Ctermini of the peptide and subsequently linked to the N-terminus of recombinantly expressed Fc using native chemical ligation. The linker length between the ANP and Fc moieties was varied among 2, 11, or 16 amino acids. In addition, either one (“monomeric”) or two (“dimeric”) ANP peptides were linked to Fc to study whether this modification had an effect on in vitro activity and/or in vivo half-life. The various constructs were studied for in vitro activity using a cell-based cGMP assay. The ANP-Fc fusion constructs were between 16- and ∼375-fold weaker than unconjugated ANP in this assay, and a trend was observed where the most potent conjugates were those with longer linkers and in the dimeric configuration. The pharmacokinetics of several constructs were assessed in rats, and the half-life of the ANP-Fc’s were found to be approximately 2 orders of magnitude longer than that of the unconjugated peptide. There was no significant difference in terminal half-life between the monomeric and dimeric constructs (2.8−5.5 h), but a trend was observed where the Cmax of the monomeric constructs was approximately 3-fold higher than that of the dimeric constructs, although the origin of this effect is not understood. These novel ANP-Fc fusion constructs hold promise for future therapeutic application in the treatment of cardiovascular diseases.



degradation.11 Subsequent shuttling and release of Fc at the cell surface serves to extend the in vivo half-life of Fc-linked moieties. The strategy of Fc fusions has been applied successfully to both peptides and proteins. An interesting adaptation of the Fc fusion approach is the generation of monovalent protein-Fc “monomers”, whereby only a single effector moiety is linked to the dimeric Fc molecule.12 This monomeric Fc fusion approach has resulted in enhanced biological activity and pharmacokinetic properties for several proteins when compared to those properties for the dimeric forms. Lastly, Fc fusion technology provides the possibility of efficient noninvasive delivery of Fc fusion proteins via FcRn.13 For example, an erythropoietin-Fc fusion protein was absorbed through the lungs with high efficiency in cynomolgus monkeys using the FcRn pathway.14 We hypothesized that the Fc domain fusion to ANP would enhance the half-life of ANP via FcRn recycling, and also provide a degree of protection against protease cleavage and NPR-C clearance by steric hindrance. Herein, we describe the semisynthetic generation of ANP-Fc monomers and dimers using recombinant Fc and synthetic peptides. The ANP-Fc fusion proteins were characterized by several analytical

INTRODUCTION Atrial natriuretic peptide, ANP, was discovered in 1981 after it was observed that injection of heart atrial muscle extracts into rats resulted in natriuresis.1 ANP was subsequently isolated in 1984 and characterized as a 28 amino acid peptide with a single cysteine disulfide loop.2 ANP is released into circulation by the cardiac muscle when the heart undergoes increased atrial stretching. The peptide binds and activates the natriuretic peptide receptor A (NPR-A, also known as GC-A receptor), which is expressed in a variety of tissues, to release intracellular cGMP and trigger further signal transduction. The end result of ANP/NPRA/cGMP signaling is increased water and salt excretion, and thus lower blood volume and blood pressure.3 Consequently, there has been intense research to develop ANP, and related natriuretic peptide factors, into viable cardioprotective drugs. One major challenge in the development of ANP-related drugs is the very short plasma half-life of ANP itself (2−5 min)4−6 caused primarily by cleavage of the peptide by neutral endopeptidase NEP 24.11 and the clearance receptor natriuretic peptide receptor C (NPR-C).7 One approach to extending the half-life of peptide8,9 and protein10 therapeutics is fusion of the potential therapeutic moiety to the Fc domain of IgG for recycling via the neonatal Fc receptor, FcRn. FcRn is broadly expressed in a variety of tissues including endothelial cells and is thought to bind Fc in acidic endosomes, thereby diverting the protein from lysosomal © 2012 American Chemical Society

Received: November 6, 2011 Revised: January 16, 2012 Published: January 22, 2012 518

dx.doi.org/10.1021/bc200592c | Bioconjugate Chem. 2012, 23, 518−526

Bioconjugate Chemistry

Article

methods, as well as by an in vitro cGMP induction assay. Finally, the pharmacokinetics of the ANP-Fc fusions were assessed in rats for two monomeric and dimeric constructs.

Fractions containing >90% peptide were pooled, flash frozen, and lyophilized. Synthesis of N-Terminal ANP Thioesters. After automated synthesis, the N-terminus of the peptide thioesters was converted to a free carboxylic acid by manually treating the resin with succinic anhydride (10 equiv) and DIEA (10 equiv) in DMF for 2 h at RT. Reaction contents were removed via vacuum filtration, and the resin was washed with DMF and DCM, respectively. To generate the thioester, this material was treated with 5 equiv of Gly(SBn)·HCl, 5 equiv of PyBOP (Novabiochem), and 15 equiv of DIEA in DMF and stirred for 18 h at RT. The protected peptide thioester was cleaved from the resin and deprotected as described above. The peptide was isolated and purified as described above. Conjugation of ANP Thioester Peptides to Cys-Fc. Cys-Fc (5 mg/mL in PBS, pH 7.4) was treated with 2mercaptoethanesulfonic acid, sodium salt (MESNA, SigmaAldrich), such that the final concentration of MESNA was 20 mM. Peptide thioester was added to the reaction mixture (2−6 mol equiv depending on the peptide) and allowed to mix for 18 h at RT. The actual number of equivalents of peptide was selected based on pilot scale reactions such that an equal percentage of monomer (1 ANP peptide per Fc) and dimer (2 ANP peptides per Fc) conjugate was formed. The crude reaction mixture was diluted to 1 mg/mL in PBS and dialyzed against PBS (pH 7.4) (7 changes over 24 h). NuPage SDS-PAGE gels (Invitrogen) were used to determine the extent of reaction. Prior to protein purification, the conjugate was dialyzed into 50 mM Tris-HCl buffer pH 7.2. Purification of the Semisynthetic ANP-Fc Conjugates. The dialyzed ANP-Fc conjugation reaction mixture was adjusted to a final concentration of 50 mM sodium acetate, pH 5.0, with 1 M sodium acetate, pH 5.0, then filtered through a 0.8/0.2 μm syringe filter. This clarified solution was loaded onto a 1 × 10 cm column packed with Fractoprep SO3-650(M) cation exchange resin (CEX) equilibrated with 50 mM sodium acetate, pH 5.5. After loading the column was washed with 3 column volumes (CV) of equilibration buffer. The protein was eluted with a linear gradient from 0 to 0.5 M sodium chloride (in 50 mM sodium acetate pH 5.5) over 30 CV. Fractions were analyzed using NuPage SDS-PAGE gels and the fractions containing the majority of the monomer conjugate were pooled together, while separately the fractions containing the majority of the dimer conjugate were pooled together. The CEX pools were first adjusted to 0.1 M Tris pH 7.2 using a 1 M Tris pH 7.2 stock solution and were then adjusted to 1 M ammonium sulfate using a 3 M ammonium sulfate stock solution. After filtration with a 0.8/0.2 μM syringe filter, this clarified protein solution was loaded onto a 1 × 10 cm column packed with EMD Fractogel TA 650(S) thio-affinity column pre-equilibrated with 50 mM Tris, 1 M Ammonium Sulfate, pH 7.2. After washing the column with 3 CV of equilibration buffer, the protein was eluted with a decreasing linear gradient from 1.0 to 0.33 M ammonium sulfate over 20 CV. This was followed by a step gradient from 0.33 to 0 M Ammonium Sulfate. The peak fractions containing highly purified conjugate were identified using SDS-PAGE gels then pooled together. The final pool was dialyzed against 1× PBS. Protein concentration was determined using UV-280 nm analysis. If needed, the final protein was concentrated using Amicon Ultra-15 Centrifugal Concentration Unit(s) [Millipore]. The final protein was aseptically filtered through a 0.2 μm filter then aliquots were stored at −80 °C.



MATERIALS AND METHODS Cloning, Expression, and Purification of Cys-Fc. The human Cys-Fc sequence from IgG1 was obtained from a human leukocyte cDNA library with the mouse Igκ signal sequence using standard PCR techniques. The Cys-Fc sequence was then cloned into pEE12.4 (Lonza Group) to create pSYN-CysFc-004. See Supporting Information for protein sequence of CysFc. CHOK1SV suspension cells (Lonza) were transfected with pSYN-cysFc-004 by electroporation according to the manufacturer’s instructions. Cells were selected with 25 μM MSX at 37 °C/5% CO2 for three weeks. Transfected cells were expanded and adapted to serumfree suspension culture. CysFc expression levels were assessed using a protein A immunoprecipitation procedure followed by nonreducing SDS-PAGE analysis. Based on immunoprecipitation results, the cell line 9F4 was chosen for production. Cys-Fc was produced from 20 L of 9F4 cells. At the end of the production run, cells were removed by centrifugation and the conditioned medium concentrated 4-fold by tangential flow filtration. Cys-Fc was purified by protein A chromatography and eluted with 0.1 M sodium citrate containing 0.15 M sodium chloride, pH 3.4. The protein was then diafiltered into DPBS and stored at 4 °C until use. Peptide Synthesis. Peptides were synthesized on an Advanced ChemTech 396Ω synthesizer using Fmoc-GlyNovaSyn TGT resin (Novabiochem) and standard Fmocsolid phase peptide synthesis using HBTU as the coupling agent, N,N-diisopropylethylamine (DIEA) as the base and N,Ndimethylformamide (DMF) as the solvent. Crude peptides were purified by reverse phase HPLC (Phenomenex Jupiter 5 μ C4 300 Å column, 250 × 21.20 mm) using gradients of acetonitirile in water with 0.1% TFA. Peptide purity was confirmed >90% by analytical RP-HPLC, and peptide identities were confirmed by electrospray MS. Synthesis of C-Terminal ANP Thioesters. After automated synthesis, the N-terminus was protected by manually treating the resin with 10 equiv of di-t-butyl dicarbonate (Boc2O) and 20 equiv of DIEA in DMF, and allowed to mix overnight at RT. The fully protected peptide was cleaved from the resin by mild acid cleavage in DCM with 30% 1,1,1,3,3,3hexafluoro-2-propanol (HFIP) for two hours. The resin was filtered and the filtrate was concentrated in vacuo overnight, yielding the protected peptide with a free C-terminus. This material was treated with 1.5 equiv of glycine benzyl thioester hydrochloride salt (Gly(SBn)·HCl), 1.5 equiv of PyBOP (Novabiochem), and 4.5 equiv of DIEA in DMF and stirred for 18 h. The reaction mixture was concentrated in vacuo for 18 h. The protected peptide thioester was deprotected by treatment with 25 mL of cleavage cocktail (95% trifluoroacetic acid (TFA), 2.5% triisopropylsilane (TIS), and 2.5% ethanedithiol (EDT)) for two hours and 45 min, after which 2.5% (v/v) bromotrimethylsilane (TMSBr) was added and allowed to mix for 15 min. The reaction mixture was concentrated, and the crude peptide thioester was precipitated with cold diethyl ether (Et2O). Peptide was centrifuged, supernatant decanted, and crude peptide was triturated two more times with cold Et2O. The peptide was purified by HPLC, and fractions were analyzed by LC/MS for identity and purity. 519

dx.doi.org/10.1021/bc200592c | Bioconjugate Chem. 2012, 23, 518−526

Bioconjugate Chemistry

Article

Size-Exclusion Chromatography of the Semisynthetic ANP-Fc Fusions. The aggregation content of the semisynthetic ANP-Fc fusion proteins was assessed using analytical size exclusion chromatography (SE-HPLC). A TSKgel Super SW2000 4.6 mm × 300 mm (Tosoh Biosciences) SE-HPLC column was run at a flow rate of 0.4 mL/min using PBS, pH 7.4, and 250 mM NaCl as the eluent. Samples of 5 μL of a 1.0 mg/mL solution were injected for each analysis. BioRad MW standards were injected prior to and after each set of injections to ensure the integrity of the column. Area under the curve integration was performed to determine the percent of each separated component peak. Characterization of Free Peptide in the Semisynthetic ANP-Fc Fusions. The free unconjugated ANP peptide contaminant levels in the purified semisynthetic ANP-Fc fusions preps were analyzed using reversed-phase chromatography. A Protein C4 reversed phase column (Grace Vydac) was injected with 100 μg of the purified semisynthetic Fc fusion protein. The two mobile phases were A) 0.1% TFA in water and B) 0.1% TFA in acetonitrile. The reversed phase chromatography was run using the following gradient profile over a total run time of 50 min (min): 0−0.5 min at 5% B, 5− 35 min ramp up to 95%B, 35−40 min hold at 95%B, 40−42 min ramp down to 5%B, 42−50 min hold at 5%B. ANP fusions were analyzed neat and spiked with 0.1 μg of unconjugated free ANP peptide (0.1 μg spike is equivalent to 1.3 mol %) to visualize expected retention time and peak heights of potential contamination. The limit of quantitation of unconjugated free ANP using this method was 1.3% (mole/mole) (potential ANP contamination level). In addition, a 10 μg injection of the corresponding unconjugated peptide was injected to verify the retention time of the potential free peptide contamination. Characterization of Peptide Disulfide Linkages in the Semisynthetic ANP-Fc Fusions. The disulfide linkages of the ANP-Fc fusion monomers and dimers were identified by combining trypsin digestion and nano-LC/MS/MS under reducing and nonreducing conditions (see Supporting Information). An Agilent 1100 HPLC binary system and a Thermo LCQ-Advantage ion trap mass spectrometer were used for the LC/MS/MS analysis. The unconjugated ANP peptide analogue (including GS linker region) of the ANP-Fc fusions was used as the positive control for the identification experiment. A 3 μg sample of peptide-Fc fusion (1 μg of free peptide in oxidized form) was diluted to a final volume of 30 μL with 100 mM ammonium acetate (pH 5.25). A 10 μL sample of a 0.01 μg/mL trypsin solution (trypsin gold, MS grade, Promega) was added to the protein/peptide solution and the digestion was carried out at 37 °C for 18 h. The digested sample was split into two aliquots of 20 μL. One aliquot of the digest (nonreduced condition) was frozen immediately and stored at −20 °C until analysis. The second aliquot was dried down in vacuo and redissolved with 20 μL of 50 mM Tris-HCl and 1 mM EDTA (pH 8.0). One microliter of DTT (0.2 M) was added and the solution was incubated at 56 °C for 30 min. Subsequently, 1 μL of 2-iodoacetamide (0.5 M) was added and the reaction proceeded for 30 min in the dark to generate the thiol-capped digestion sample. The unconjugated peptide digest was diluted 5-fold with water, and 0.5 μL of digested peptide-Fc fusion (or 1 μL of digested peptide) was loaded onto a self-packed C18 capillary LC column (YMC C18, 5 μm, 10 cm long) for nano-LC/MS/MS analysis. The HPLC pump flow rate of 0.2 mL/min was split to obtain 1 μL/min for the capillary LC. The mobile phases used were 0.05 M acetic acid

in water (A) and 0.05 M acetic acid in acetonitrile (B). The LC/MS data (nonreduced and reduced) were compared to theoretical masses corresponding to possible disulfide folding linkages for the digested fragments to determine whether or not the proper disulfides within the ANP peptide(s) had formed. The reduced peptide fragment sequence was confirmed by SEQUEST database searching of MS/MS data. The peptide digest data also confirmed the conjugation of the peptides to the N-termini of CysFc. Characterization of the FcRn Binding Properties of the Semisynthetic ANP-Fc Fusions Using Surface Plasmon Resonance. All reagents for immobilization of FcRn to the Biacore chip were purchased from Biacore AB (Uppsala, Sweden). Soluble human FcRn (produced in CHO cells) was cross-linked to the dextran surface of a CM5 sensor chip by amine coupling using 1-ethyl-(3-dimethylaminopropyl)-carbodiimide hydrochloride (EDC) and N-hydroxysuccinimide (NHS) as recommended by Biacore. For immobilization, the FcRn was diluted to 10 μg/mL in 10 mM sodium acetate, pH 4.5. Residual sites on the dextran were blocked with 1 M ethanolamine hydrochloride pH 8.5. FcRn was immobilized to one flow cell on the sensor chip, while a control flow cell was blocked with ethanolamine for reference subtraction. Sufficient FcRn was bound to the chip to result in 400 to 600 response units (RU). Experiments were performed at pH 6 using 50 mM sodium phosphate, 100 mM sodium chloride, and 0.01% surfactant P20 (Biacore AB). Eleven 2-fold serial dilutions between 1 μM and 1 nM of the ANP-Fc or control proteins were injected over the FcRn-CM5 chip at 30 μL/min for 5 min. Bound protein was then dissociated from the chip for 2.5 min with running buffer. Any remaining protein was removed from the chip with a 27 s injection of pH 7.4 HBS-P buffer (0.01 M HEPES pH 7.4, 0.15 M NaCl, 0.005% v/v Surfactant P20) at 30 μL/min. Sensorgrams were analyzed using BiaEval software version 3.1 (Biacore AB) and were baseline corrected using a buffer blank then baseline averaged. The Response Units at equilibrium (Req) values are plotted against concentration and the equilibrium affinity constants (KD’s) are then derived by fitting the data to a heterogeneous ligand model using BiaEval software. This model assumes that there are two classes of noninteracting binding sites of FcRn on the sensor chip. Thus, the KD values are reported as two separate KD’s that contribute some fractional % to the theoretical maximal observed Req value (Rmax).15,16 Generation of Stable Cell Lines for In Vitro NPRA Cell Assays. Full-length human NPRA, plasmids were purchased from OriGene Technologies, Inc., or sequence synthesized and subcloned into pcDNA3.1 mammalian expression vector (Invitrogen). The pcDNA3.1 NPR clones were transfected using Lipofectamine (Invitrogen) into HEK293 cells and stable cell lines were selected using G418. Clones were screened using the ANP-induced cGMP assay described below. High cGMP producing clones were expanded in DMEM containing, 100 μg/mL penicillin/streptomycin, L-glutamine, 400 μg/mL of G418, and 10% FBS (Hyclone, Logan, UT). Natriuretic Peptide Induced cGMP Assay. HEK293 NPRA cells grown to 90% confluence were harvested with Versene (Invitrogen). Cells were washed and resuspended at 3.3 × 105 cells per mL in prewarmed Dulbecco’s PBS, pH 7.4, 25 mM HEPES, 0.1% BSA, 1 mM 3-Isobutyl-1-methylxanthine (IBMX) [Assay Buffer]. 40 μL of cell suspension was added to 40 μL of ANP or ANP-Fc in Assay Buffer in triplicate and 520

dx.doi.org/10.1021/bc200592c | Bioconjugate Chem. 2012, 23, 518−526

Bioconjugate Chemistry

Article

Figure 1. Schematic representation of the semisynthetic method used to generate the ANP-Fc fusion proteins.

conjugated antibody (Pierce Biotechnology, cat #31416) diluted 1:25 000 in PBS/2%BSA for 1 h at RT. Plates were washed three times with 300 μL/well of PBST before development with 100 μL/well of TMB solution (1:1 mix of TMB substrate and H2O2; Pierce) and until color developed. Reactions were stopped by addition of 2 M sulfuric acid (100 μL/well). Plates were read at 450 nm in a Spectromax plate reader.

incubated for 20 min at 37 °C. The cGMP concentration was measured using the HitHunter cGMP Assay Kit (DiscoveRx, Corporation Freemont, CA). cGMP production dose response curves were generated with a four parameter logistic equation fitted using the Levenberg−Marquardt algorithm in XLf it4.2 data analysis software (ID Business Solutions, Ltd., Guildford, UK). Pharmacokinetics of ANP-Fc 2m, 2d, 3m, and 3d. Female Wistar rats (∼100 g; 4 rats/protein) were dosed IV with 0.5 mg/kg protein 2m, 2d, 3m, and 3d in PBS. Blood was collected from each rat by tail nick at 1, 2, 4, 8, 24, 48, and 72 h (protein 3d) or 0.25, 1, 2, 4, 8, 24, 48, and 72 h (proteins 2m, 3d, and 3m) after dosing. Blood (2 × 60 μL aliquots) was collected into microcapillary tubes (Fisher Scientific, catalog #22−362−574) containing one-tenth volume of 3.2% sodium citrate and plasma was generated by centrifugation. Plasma was stored at −20 °C until analysis by ANP-Fc/Fc ELISA and Fc/ Fc ELISA. Pharmacokinetic parameters were assessed using WinNonlin pharmacokinetic software (Pharsight v 4.1 or 5.1) All studies in animals were conducted using protocols that were approved by an Institutional Animal Care and Use Committee (IACUC), following the National Institutes of Health guidelines for the care and use of research animals. ANP-Fc/Fc ELISA. 96-well plates (Costar, catalog #3369) were coated with 5 μg/mL (50 μL/well) mouse antihuman ANP (US Biologicals, catalog #A4150) in 50 mM of carbonate/ bicarbonate buffer pH 9.6 at 4 °C overnight. Plates were blocked with 300 μL/well of PBS/5% BSA (Jackson ImmunoResearch, catalog #001−000−162) for 2 h at RT. Standards and samples (100 μL/well) were diluted in PBS/5% BSA and incubated for 2 h at RT. Standard curves ranged from 0.039 ng/mL to 10 ng/mL. Plates were washed three times with 300 μL/well of PBS containing 0.05% Tween-20 (PBST) in a Tecan plate washer. Plates were then incubated with 100 μL/well of goat antihuman (Fc specific) horseradish peroxidase (HRP) conjugated antibody (Pierce Biotechnology, catalog #31413) diluted 1:7500 in PBS/5%BSA for 3 h at RT. Plates were washed four times with 300 μL/well of PBST before development with 100 μL/well of TMB supersensitive substrate (BioFx Laboratories) at RT for approximately 6 min. Reactions were stopped by addition of 0.25 M of sulfuric acid (100 μL/well). Plates were read at 450 (−600) nm in a Spectromax plate reader. Fc/Fc ELISA. 96-well plates (Costar, catalog #3369) were coated with 1 μg/mL (50 μL/well) goat antihuman IgG, Fc fragment (Pierce, catalogue #31123, stock = 1.8 mg/mL) in 0.1 M carbonate/bicarbonate buffer pH 9.2 and incubated at 37 °C for 1 h. Plates were blocked with 200 μL/well of PBS/2% BSA for 1 h at 37 °C. Standards and samples (100 μL/well) were diluted in PBS/2% BSA and incubated for 1 h at 37 °C. Plates were washed three times with 300 μL/well of PBST in a Tecan plate washer. Plates were then incubated with 100 μL/well of goat antihuman (Fc specific) horseradish peroxidase (HRP)



RESULTS Generation of Semisynthetic ANP-Fc Constructs. The ANP-Fc proteins were generated semisynthetically using native ligation chemistry.17 This process involved the reaction of a recombinantly produced Fc fragment of IgG1 containing a Nterminal cysteine (CysFc) with various synthetically produced ANP thioesters to generate a native amide bond at the linkage site (Figure 1). ANP peptide was synthesized with the thioester at either the N-terminus or the C-terminus of ANP to study the effect of the different Fc fusion orientations on activity. To generate the CysFc protein, Fc from IgG1 was truncated at the first naturally occurring N-terminal hinge cysteine of Fc (NH2− CPPCDKTH...; IgG1) such that no new cysteines were introduced into the sequence of Fc, yet retaining both hinge cysteines for hinge disulfide formation. The linker connecting ANP and Fc was composed of glycine and serine residues (GGS) to generate a flexible hydrophilic linker in the hopes of having minimal interference with the biological activity of ANP. Although additional non-natural linkers such as PEG would have been possible to incorporate into the model system, natural linker residues were selected in this study to model recombinant protein production. In addition, the linkers were designed to be stable and noncleavable for this initial study; however, future designs could also incorporate cleavable linker/prodrug technology in efforts to boost overall activity.18,19 Lastly, the linker length was also varied from 2, 11, and 16 amino acids (Figure 2) to probe the effect of linker length on activity. The maximum length of a 16 amino acid linker was driven by peptide availability for this study; linkers of >16 amino acids would also be of interest to study in the future. CysFc was cloned, expressed in CHO cells, and purified by protein A chromatography. Binding of CysFc to FcRn was confirmed using surface plasmon resonance (SPR) analysis (Table 2, vide inf ra). ANP thioesters were synthesized using standard Fmoc/tBu protocols with a few postsynthesis modifications. In the case of C-terminal thioesters, a mild acid-sensitive trityl-based polystyrene resin was used. After synthesis of the peptide on resin, treatment of the protected peptide with 1,1,1,3,3,3-hexafluoro-2-propanol in dicholoromethane cleaved the peptide from the resin, but left all of the remaining protecting groups intact. The free C-terminus of the peptide was coupled to glycine benzyl thioester (NH2-GlySBn), then treated with TFA to cleave the remaining protecting 521

dx.doi.org/10.1021/bc200592c | Bioconjugate Chem. 2012, 23, 518−526

Bioconjugate Chemistry

Article

Figure 2. Schematic illustrating the synthesis and structures of various ANP peptides and the structures of various ANP-Fc protein constructs. Nomenclature: ‘p’, peptide; ‘m’, ANP-Fc monomer; ‘d’, ANP-Fc dimer.

groups (Figure 2).20 In the case of N-terminal thioesters, after peptide synthesis, the resin was treated with succinic anhydride to generate an N-terminal carboxylic acid which was subsequently coupled to NH2-Gly-SBn to generate the N-terminal thioester. Subsequent purification by reversed phase HPLC provided peptides 1p−5p (Figure 2). Ligation of the ANP thioesters to CysFc was performed at neutral pH in the presence of a reducing agent and using a stoichiometrically limiting amount of peptide thioester, such that the reaction would generate a mixture of ANP-Fc “monomers” (single peptide linked to Fc), and ANP-Fc “dimers” (two peptides linked to Fc). The protein reaction mixtures were then extensively dialyzed into 50 mM Tris-HCl buffer at pH 7.2 to

ensure that the hinge disulfide at the N-termini was regenerated as initially assessed by nonreducing SDS-PAGE. The proteins were purified using a two-step process: cation exchange chromatography followed by thiol-affinity chromatography. The cation exchange step separates the CysFc, the monomer and dimer ANP-Fc conjugates, while also removing the protein aggregates. The individual ANP-Fc conjugate cation exchange pools were independently put through a thiol-affinity chromatography purification step to remove the unconjugated free ANP peptide and to further purify the monomer and dimer conjugate species. It should be noted that the thiol-affinity method was developed and utilized as a result of its clean separation of the protein conjugates and the free peptide. 522

dx.doi.org/10.1021/bc200592c | Bioconjugate Chem. 2012, 23, 518−526

Bioconjugate Chemistry

Article

Therefore, unconjugated peptide is not expected to be present in the final purified proteins. Analytical Characterization of Semisynthetic ANP-Fc Proteins. The purified ANP-Fc proteins were characterized by SDS-PAGE which was able to differentiate the proteins with zero, one, and two ANP peptides linked to Fc, with bands migrating consistent with their expected molecular weight (Figure 3). Densitometry was performed on the gels and it was

reaction (Table 1 and Supporting Information). No ANP peptide was detected in any of the protein preparations where the limit of detection for this method was 1.3% (mol/mol) free ANP peptide. Finally, trypsin digestion/LC/MS/MS analysis was performed to ensure that the proper disulfide patterns within both ANP and Fc were present, and to confirm the site of conjugation to CysFc. The semisynthetic constructs possessed the expected ANP disulfide folding pattern, although a small fraction of disulfide misfolding was observed for some of the constructs. The percentage of misfolded disulfides was estimated to be 3−15% of the total (see Supporting Information), therefore the proteins were considered acceptable for further analysis. The trypsin digest also confirmed that site of peptide conjugation was at the N-terminal cysteine of CysFc as expected based on the selectivity of the native ligation methodology. Affinity of Semisynthetic ANP-Fc Proteins to Human FcRn by SPR. The affinity of each ANP-Fc construct for human FcRn was assessed using surface plasmon resonance (SPR) to verify that the synthetic chemistry did not alter the FcRn-binding capabilities of the Fc domain. No significant changes in FcRn binding were observed as compared to the Cys-Fc control (Table 2 and Supporting Information).

Figure 3. Representative SDS-PAGE analysis of ANP-Fc monomer 3m and dimer 3d. Lanes 1−4: nonreducing conditions; Lanes 5−8: reducing conditions. Lanes 1,5: molecular weight markers. Lanes 2,6: CysFc. Lanes 3,7: protein 3m. Lanes 4,8: protein 3d.

Table 2. Summary of FcRn Surface Plasmon Resonance Data for Each of the Semisynthetic ANP-Fc Proteinsb

estimated that each protein was >88% pure (Table 1). Analytical size-exclusion chromatography was used to monitor Table 1. Summary of Analytical Characterization of ANP-Fc Fusion Proteins by SDS-PAGE, SEC-HPLC, and RP-HPLCa protein name

reducing SDS-PAGE (% Purity)b

non-reducing SDS-PAGE (% Purity)

SE-HPLC % Expected Major Peak

RP-HPLCc % (mol/mol) unconjugated peptide

1m 1d 2m 2d 3m 3d 4m 4d 5m 5d

48 + 40 = 88 99 53 + 44 = 97 99 59 + 37 = 96 99 53 + 43 = 96 99 53 + 39 = 92 99

93 92 97 95 96 97 99 95 98 99

97 99 99 99 >95 n/a 99 95 96 99