Recombinant Protein Hydrazides - American Chemical Society

May 14, 2011 - mediated cleavage of the corresponding intein fusion protein.7,8 .... washed beads (lane 7), cleaved and washed chitin beads (lane 8). ...
0 downloads 0 Views 2MB Size
COMMUNICATION pubs.acs.org/bc

Recombinant Protein Hydrazides: Application to Site-Specific Protein PEGylation Jennifer Thom,† David Anderson,† Joanne McGregor,†,‡ and Graham Cotton*,† †

Almac Sciences, Elvingston Sciences Centre, Gladsmuir, East Lothian, EH33 1EH, United Kingdom

bS Supporting Information ABSTRACT: Here, we describe a novel method for the sitespecific C-terminal PEGylation of recombinant proteins. This general approach exploits chemical cleavage of precursor inteinfusion proteins with hydrazine to directly produce recombinant protein hydrazides. This unique functionality within the protein sequence then facilitates site-specific C-terminal modification by hydrazone-forming ligation reactions. This approach was used to generate folded, site-specifically C-terminal PEGylated IFNalpha2b and IFNbeta1b, which retained excellent antiviral activity, demonstrating the utility of this technology in the PEGylation of therapeutic proteins. As this methodology is straightforward to perform, is compatible with disulfide bonds, and is exclusively selective for the protein C-terminus, it shows great potential as general technology for the site-specific engineering and labeling of recombinant proteins.

R

ecombinant protein therapeutics have emerged as an effective treatment for a variety of conditions ranging from cancer to metabolic disorders and autoimmune diseases, but they are commonly limited by their poor pharmacokinetics and immunogenicity. Arguably, the most widely used and accepted method for improving the pharmacokinetics of proteins is through PEGylation.1,2 The covalent attachment of poly(ethylene glycol) (PEG) increases the half-life of the protein in vivo by decreasing renal clearance, due to the large hydrodynamic radius of the polymer, as well as masking potentially immunogenic epitopes and protease cleavage sites. Furthermore, PEGylation of a target protein can substantially improve its solubility and stability. There are currently nine PEGylated protein therapeutics approved for clinical use;1 however, established methods for PEGylation are generally non-site-selective, resulting in a mixture of protein PEG-positional isomers with significantly decreased biological activity. For example, ViraferonPEG/PegIntron (Schering Plough), approved for the treatment of Hepatitis C, is produced by conjugating recombinant interferon alpha2b (IFNalpha2b) to a single chain 12 kDa succinimidyl carbonate PEG. This results in 95% monoPEGylated protein comprising 14 positional isomers, and as a consequence, PegIntron retains only 28% antiviral activity in in vitro assays compared to non-PEGylated IFNalpha2b.3 There is therefore a pressing need for methods to sitespecifically PEGylate proteins. By incorporating a single PEG moiety at a defined position within the protein sequence, the deleterious effects associated with nonselective PEGylation may be minimized. In addition, approaches that yield homogeneous PEGylated proteins are potentially attractive from a regulatory perspective. While a number of protein engineering approaches have been applied to this end,46 one particularly versatile technology for site-specific protein modification is expressed r 2011 American Chemical Society

protein ligation (EPL). The basis of EPL is the production of recombinant C-terminal thioester proteins through thiolmediated cleavage of the corresponding intein fusion protein.7,8 The protein thioester functionality enables peptides and labels to be chemoselectively attached to the protein using the wellestablished native chemical ligation reaction.9 In addition, C-terminal cleavage of the protein thioester with nucleophiles has also been used to introduce reactive moieties into proteins to enable subsequent modification through alternative chemistries.1012 While the EPL approach has proven extremely powerful, the steps are performed in the presence of thiol additives, which may compromise its use with certain protein families (for example, disulfide bond containing proteins), and the protein thioesters themselves are often labile. Here, we report the development of a versatile complementary protein ligation approach for site-specific protein PEGylation, which is compatible with folded disulfide bond-containing proteins. This process is based on chemical interception of intein-mediated protein splicing,13 to create C-terminal hydrazide recombinant proteins. This facile production of recombinant hydrazide proteins then facilitates direct site-specific C-terminal PEGylation (or other modifications) of the target protein using hydrazone bond-forming reactions14,15 (Figure 1). In principle, any PEG derivative (or alternative label) containing the appropriate functionality, namely, a ketone or aldehyde, can be chemoselectively attached to the C-terminus of the protein in this one-step process. For the approach described here, we have developed novel pyruvoyl PEG derivatives for attachment to the

Received: March 18, 2011 Revised: May 11, 2011 Published: May 14, 2011 1017

dx.doi.org/10.1021/bc2001374 | Bioconjugate Chem. 2011, 22, 1017–1020

Bioconjugate Chemistry

Figure 1. Use of intein technology to generate recombinant C-terminal hydrazide proteins for site-specific PEGylation. (a) The recombinant protein of interest (IFNalpha2b shown here) is genetically fused to the N-terminus of an engineered GyrA intein and chitin binding domain (CBD), expressed in E. coli and purified by immobilization onto chitin beads. (b) The first residue of this intein domain is a cysteine. The properties of the intein domain are such that it induces an N to S acyl shift at this protein-intein junction to form a branched thioester intermediate. (c) This thioester intermediate is chemically cleaved using hydrazine to liberate the corresponding C-terminal hydrazide derivative of the target protein. The intein-CBD remains bound to the chitin beads. (d) The protein C-terminal hydrazide chemoselectively reacts with ketone or aldehyde containing moieties. Here the reaction with a pyruvoyl functionalized PEG is shown, resulting in site-specific C-terminal PEGylation of the target protein via the resonance stabilized R-oxo hydrazone linkage.

protein C-terminus, as the resulting R-oxo hydrazone bonds are stabilized through resonance effects.16,17 IFNalpha2b was chosen as our initial target for this C-terminal PEGylation approach. As discussed above, this therapeutically important protein is currently administered as the PEG isomer mixture, and thus there is scope for improvement through sitespecific PEGylation. IFNalpha2b was expressed in E. coli as a soluble N-terminal fusion to the Mycobacterium xenopi GyrA intein, and purified from the supernatant by immobilization on chitin beads using a chitin binding domain (CBD) fused to the C-terminus of the intein domain (Figure 2a lane 3 and Supporting Information). The immobilized IFNalpha2b-intein-CBD protein was cleaved with hydrazine to generate the corresponding IFNalpha2b C-terminal hydrazide (Figure 2a,b). The isolated protein (Figure 2c lane 2) contained two disulfide bonds (Supporting Information) and had potent antiviral activity (Figure 2d) indicating that folded IFNalpha2b C-terminal hydrazide was directly produced after expression and hydrazine cleavage of the corresponding intein fusion protein.

COMMUNICATION

Figure 2. Generation and antiviral activity of site-specifically C-terminal PEGylated IFNalpha2b. (a) SDS PAGE analysis of IFNalpha2b-inteinCBD fusion protein expression in E. coli, purification using chitin beads and hydrazine cleavage. Protein standard (lane 1), cell lysate (lane 2), IFNalpha2b-intein-CBD purified by immobilization on chitin beads (lane 3), supernatant from hydrazine cleavage (lane 4), supernatant from washed beads (lane 5), cleaved and washed chitin beads (lane 6). Arrows indicate IFNalpha2b-intein-CBD fusion protein (i), intein-CBD (ii), and IFNalpha2b C-terminal hydrazide (iii). (b) Electrospray mass spectra of purified IFNalpha2b hydrazide, expected mass (without the N-terminal methionine) is 19 340 Da. The higher observed mass probably reflects methionine oxidation, previously reported for IFNalpha2b and shown not to significantly affect the activity of the protein.24 (c) SDS PAGE analysis of IFNalpha2b hydrazide (lane 2) and purified C-terminal PEGylated IFNalpha2b (lane 3). (d) Antiviral activity of IFNalpha2b derivatives in a cytopathic effect inhibition assay using human A549 lung carcinoma cells and EMC virus. Measured activities are in million international units, MIU/mg IFNalpha2b protein ( SD where appropriate; IFNalpha2b hydrazide and C-terminal PEGylated IFNalpha2b (n = 8), IFNalpha2b standard (n = 2) and ViraferonPEG (n = 2). For reported values, see Supporting Information. IFNalpha2b standard and ViraferonPEG activities are in line with previous reports.3.

This is in contrast to standard intein cleavage methods that use thiols such as dithiothreitol (DTT) and sodium mercaptoethylsulfonate (MESNa) that usually reduce disulfide bonds within the target protein. For the PEGylation reactions, a pyruvoyl derivative of 10 kDa PEG was synthesized and incubated with IFNalpha2b hydrazide as described in the Supporting Information. This typically afforded 6075% yield of the C-terminal PEGylated IFNalpha2b protein (Supporting Information, Figure S1). Importantly, no PEGylation product was observed when a C-terminal thioester derivative of IFNalpha2b was incubated with pyruvoyl PEG under the same conditions, consistent with site-specific PEGylation via the C-terminal hydrazide group only (data not shown). PEGylated IFNalpha2b was purified in 2 steps: ion exchange to remove unreacted pyruvoyl-PEG, followed by gel filtration to separate PEGylated IFNalpha2b from any unreacted IFNalpha2b hydrazide (Supporting Information, Figure S2) to yield pure C-terminal, mono PEGylated IFNalpha2b (Figure 2c, lane 3). The antiviral activities of the C-terminally PEGylated IFNalpha2b and IFNalpha2b hydrazide control purified under the 1018

dx.doi.org/10.1021/bc2001374 |Bioconjugate Chem. 2011, 22, 1017–1020

Bioconjugate Chemistry same conditions were measured in a cytopathic effect inhibition assay with human A549 lung carcinoma cells and encephalomyocarditis (EMC) virus (Figure 2d). The site-specifically C-terminal PEGylated IFNalpha2b retained 74% of the activity of the standard IFNalpha2b protein, more than double that of the current heterogeneous PEGylated IFNalpha2b therapeutic ViraferonPEG, clearly demonstrating the advantage of this technology. To further exemplify this site-specific C-terminal PEGylation technology, IFNbeta1b was deemed to be an attractive target. This protein is also an important therapeutic, but currently there are no PEGylated versions approved. Betaseron (Betaferon in the EU, Bayer Shering Pharma) and more recently Extavia (Novartis) are recombinant IFNbeta1b proteins used in the treatment of multiple sclerosis, and are under clinical evaluation for the treatment of other diseases including hepatitis and certain cancers.18,19 IFNbeta1b is rapidly cleared from blood necessitating frequent administration regimes that can result in injection site necrosis and decreased patient compliance. Neutralizing antibodies are also problematic, with up to 35% of patients developing these in one reported study.20 In addition, IFNbeta1b is a very unstable protein, prone to aggregation.21 Work to address these issues includes random PEGylation of side-chain primary amines, which results in a mixture of 5 major positional isomers, and PEGylation of (predominantly) the N-terminal amine, resulting in products with 2231% and 1966% activity of the unmodified IFNbeta1b, respectively.22 Attempts at site specific PEGylation using an engineered free cysteine residue at either position 79 (the natural glycosylation site) or the N- or C-termini were unsuccessful due to insufficient site specific attachment.22 Another strategy used bicin (bis-N-2-hydroxyethylglycinamide) linkers to randomly attach 23 PEG molecules. Under physiological conditions, rapid hydrolysis of the bicin releases the PEG to leave the active IFNbeta1b. However, the activity of these releasable PEGylated forms was only between 7% and 27% that of the unmodified IFNbeta1b.23 We expressed soluble IFNbeta1b-intein-CBD fusion protein in E. coli and purified the target protein through immobilization onto chitin beads. Intein cleavage was induced by overnight treatment with hydrazine and the resulting IFNbeta1b-hydrazide was purified by gel filtration (Figure 3a,b; Supporting Information). As with IFNalpha2b, IFNbeta1b C-terminal hydrazide was generated with the natural disulfide bond intact indicating that the protein was correctly folded (Figure 3a and Supporting Information). To generate site-specifically C-terminal PEGylated IFNbeta1b, IFNbeta1b C-terminal hydrazide was mixed with the pyruvoyl PEG 10 kDa derivative and left overnight at room temperature. This typically gave ∼55% yield of the C-terminal PEGylated protein (Supporting Information, Figure S3a). No PEGylation product was observed when IFNbeta1b C-terminal acid was incubated with pyruvoyl PEG under similar conditions, consistent with sitespecific PEGylation via the C-terminal hydrazide group only (Supporting Information, Figure S3b). The site-specifically C-terminal PEGylated IFNbeta1b was purified by ion exchange and gel filtration as described in the Supporting Information to generate pure C-terminal PEGylated IFNbeta1b (Figure 3c lane 3). The antiviral activity was measured in a cytopathic effect inhibition assay with human A549 lung carcinoma cells and EMC virus (Figure 3d). The activity of IFNbeta1b hydrazide was lower than that reported for Betaseron, potentially due to the inherent instability of the protein and the lack of stabilizing ingredients in our formulation. Site-specifically

COMMUNICATION

Figure 3. Generation and antiviral activity of site-specifically C-terminal PEGylated IFNbeta1b. (a) SDS PAGE analysis of IFNbeta1b-inteinCBD fusion protein expression in E. coli and the subsequent purification and hydrazine cleavage. Samples were reduced with DTT after heating in loading buffer unless stated otherwise: protein standard (lanes 1 and 4), cell lysate (lane 2), IFNbeta1b-intein-CBD purified by immobilization on chitin beads (lane 3), hydrazine cleavage supernatant without DTT (lane 5), hydrazine cleavage supernatant (lane 6), supernatant from washed beads (lane 7), cleaved and washed chitin beads (lane 8). Arrows indicate IFNbeta1b-intein-CBD fusion protein (i), intein-CBD (ii), and IFNbeta1b C-terminal hydrazide (iii). There is clearly a shift in the migration of cleaved IFNbeta1b hydrazide upon DTT treatment, consistent with the native disulfide bond being intact in the liberated protein. (b) Electrospray mass spectra of purified IFNbeta1b hydrazide, expected mass 19 950 Da. The higher observed mass probably again reflects methionine oxidation. (c) SDS PAGE analysis of purified IFNbeta1b hydrazide (lane 1) and C-terminal PEGylated IFNbeta1b (lane 3); protein standard (lane 2). (d) Antiviral activity of IFNbeta1b derivatives in a cytopathic effect inhibition assay using human A549 lung carcinoma cells and EMC virus. Activities are in million international units, MIU/mg IFNbeta protein ( SD; n = 8. The reported antiviral activity of Betaseron, the current IFNbeta1b therapeutic, is also shown (Betaseron prescribing information; http://berlex.bayerhealthcare. com/html/products/pi/Betaseron_PI.pdf). WHO calibrated IFNbeta protein was used as a standard in the assays.

C-terminal PEGylated IFNbeta1b showed greater activity than the hydrazide, probably reflecting the increased protein stability brought by the PEGylation, and this activity was in line with that reported for the current unPEGylated IFNbeta1b therapeutic, Betaseron. This, to the best of our knowledge, represents the most potent PEGylated IFNbeta1b derivative generated to date. In summary, we have demonstrated the feasibility of generating active, site-specifically C-terminal PEGylated therapeutic proteins through cleavage of the corresponding soluble intein fusion proteins with hydrazine, followed by reaction of the resulting protein hydrazide with pyruvoyl functionalized PEG. Importantly, this approach is compatible with proteins containing disulfide bonds, enabling folded proteins to be directly produced as the C-terminal hydrazide derivatives in high yield, with the integrity of any disulfide bonds and tertiary structure maintained during the PEGylation reaction. While the potential of this technology has been highlighted through the generation of 1019

dx.doi.org/10.1021/bc2001374 |Bioconjugate Chem. 2011, 22, 1017–1020

Bioconjugate Chemistry highly active C-terminal PEGylated derivatives of IFNalpha2b and IFNbeta1b, we have now used this technology to C-terminally PEGylate proteins from a variety of classes, including antibody fragments, with PEGs of different molecular weights. In addition to PEGylation, this technology has broader applications in protein engineering and has been successfully used to attach an array of probes and labels, including fluorophores and synthetic peptides, onto a variety of recombinant proteins in a site-specific fashion (unpublished data). Given the ease of the process, the yields of the associated steps, and the compatibility with disulfide bond-containing proteins, this technology offers an attractive approach for the site-specific PEGylation and modification of therapeutically important proteins and is an attractive addition to the repertoire of methodologies available for protein engineering, opening up new opportunities in this field.

’ ASSOCIATED CONTENT

bS

Supporting Information. Experimental procedures and supplementary data. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Graham Cotton, Almac Sciences (Scotland) Ltd, Elvingston Science Centre, By Gladsmuir, East Lothian, EH33 1EH United Kingdom. Tel. þ44 (0)28 3839 5879, Fax. þ44 (0)18 7540 8151, E-mail [email protected]. Present Addresses ‡

Current address: GSK, 315 Cambridge Science Park, Milton Road, Cambridge, CB4 0WG, UK.

’ ACKNOWLEDGMENT We thank Moredun Research Institute (Edinburgh, UK) for MALDI mass-spectrometry analysis, Crystelle Bermand (Almac Sciences, UK) for technical advice on the synthesis of PEG derivatives, and the Scottish Executive for financial support. ’ REFERENCES (1) Veronese, F. M., and Mero, A. (2008) The impact of PEGylation on biological therapies. BioDrugs 22, 315–329. (2) Jevsevar, S., Kunstelj, M., and Porekar, V. G. (2010) PEGylation of therapeutic proteins. Biotechnol. J. 5, 113–128. (3) Wang, Y. S., Youngster, S., Grace, M., Bausch, J., Bordens, R., and Wyss, D. F. (2002) Structural and biological characterization of pegylated recombinant interferon alpha-2b and its therapeutic implications. Adv. Drug Delivery Rev. 54, 547–570. (4) Brocchini, S., Godwin, A., Balan, S., Choi, J. W., Zloh, M., and Shaunak, S. (2008) Disulfide bridge based PEGylation of proteins. Adv. Drug Delivery Rev. 60, 3–12. (5) DeFrees, S., Wang, Z. G., Xing, R., Scott, A. E., Wang, J., Zopf, D., Gouty, D. L., Sjoberg, E. R., Panneerselvam, K., Brinkman-Van der Linden, E. C., Bayer, R. J., Tarp, M. A., and Clausen, H. (2006) GlycoPEGylation of recombinant therapeutic proteins produced in Escherichia coli. Glycobiology 16, 833–843. (6) Xie, J., and Schultz, P. G. (2006) A chemical toolkit for proteins an expanded genetic code. Nat. Rev. Mol. Cell Biol. 7, 775–782. (7) Muir, T. W., Sondhi, D., and Cole, P. A. (1998) Expressed protein ligation: a general method for protein engineering. Proc. Natl. Acad. Sci. U.S.A. 95, 6705–6710.

COMMUNICATION

(8) Muir, T. W. (2001) Development and application of expressed protein ligation. Synlett. 6, 733–740. (9) Dawson, P. E., Muir, T. W., Clark-Lewis, I., and Kent, S. B. (1994) Synthesis of proteins by native chemical ligation. Science 266, 776–779. (10) Weikart, N. D., and Mootz, H. D. (2010) Generation of sitespecific and enzymatically stable conjugates of recombinant proteins with ubiquitin-like modifiers by the Cu(I)-catalyzed azide-alkyne cycloaddition. ChemBioChem 11, 774–777. (11) Yi, L., Sun, H., Wu, Y. W., Triola, G., Waldmann, H., and Goody, R. S. (2010) A highly efficient strategy for modification of proteins at the C terminus. Angew. Chem., Int. Ed. Engl. 49, 9417–9421. (12) Kalia, J., and Raines, R. T. (2006) Reactivity of intein thioesters: appending a functional group to a protein. ChemBioChem 7, 1375–1383. (13) Cheriyan, M., and Perler, F. B. (2009) Protein splicing: A versatile tool for drug discovery. Adv. Drug Delivery Rev. 61, 899–907. (14) Gaertner, H. F., Offord, R. E., Cotton, R., Timms, D., Camble, R., and Rose, K. (1994) Chemo-enzymic backbone engineering of proteins. Site-specific incorporation of synthetic peptides that mimic the 6474 disulfide loop of granulocyte colony-stimulating factor. J. Biol. Chem. 269, 7224–7230. (15) Gaertner, H. F., Rose, K., Cotton, R., Timms, D., Camble, R., and Offord, R. E. (1992) Construction of protein analogues by sitespecific condensation of unprotected fragments. Bioconjugate Chem. 3, 262–268. (16) Bourel-Bonnet, L., Pecheur, E. I., Grandjean, C., Blanpain, A., Baust, T., Melnyk, O., Hoflack, B., and Gras-Masse, H. (2005) Anchorage of synthetic peptides onto liposomes via hydrazone and alpha-oxo hydrazone bonds, preliminary functional investigations. Bioconjugate Chem. 16, 450–457. (17) Kochendoerfer, G. G., Tack, J. M., and Cressman, S. (2002) Total chemical synthesis of a 27 kDa TASP protein derived from the MscL ion channel of M. tuberculosis by ketoxime-forming ligation. Bioconjugate Chem. 13, 474–480. (18) Fine, H. A., Wen, P. Y., Robertson, M., O’Neill, A., Kowal, J., Loeffler, J. S., and Black, P. M. (1997) A phase I trial of a new recombinant human beta-interferon (BG9015) for the treatment of patients with recurrent gliomas. Clin. Cancer Res. 3, 381–387. (19) Fukutomi, T., Nakamuta, M., Fukutomi, M., Iwao, M., Watanabe, H., Hiroshige, K., Tanabe, Y., and Nawata, H. (2001) Decline of hepatitis C virus load in serum during the first 24 h after administration of interferon-beta as a predictor of the efficacy of therapy. J. Hepatol. 34, 100–107. (20) Malucchi, S., Sala, A., Gilli, F., Bottero, R., Di Sapio, A., Capobianco, M., and Bertolotto, A. (2004) Neutralizing antibodies reduce the efficacy of betaIFN during treatment of multiple sclerosis. Neurology 62, 2031–2037. (21) Filpula, D., Yang, K., Basu, A., and Wang, M. (2005) WO2005/ 084303A2 Interferon beta polymer conjugates. PCT/US2005/006575. (22) Basu, A., Yang, K., Wang, M., Liu, S., Chintala, R., Palm, T., Zhao, H., Peng, P., Wu, D., Zhang, Z., Hua, J., Hsieh, M. C., Zhou, J., Petti, G., Li, X., Janjua, A., Mendez, M., Liu, J., Longley, C., Zhang, Z., Mehlig, M., Borowski, V., Viswanathan, M., and Filpula, D. (2006) Structure-function engineering of interferon-beta-1b for improving stability, solubility, potency, immunogenicity, and pharmacokinetic properties by site-selective mono-PEGylation. Bioconjugate Chem. 17, 618–630. (23) Zhao, H., Yang, K., Martinez, A., Basu, A., Chintala, R., Liu, H. C., Janjua, A., Wang, M., and Filpula, D. (2006) Linear and branched bicin linkers for releasable PEGylation of macromolecules: controlled release in vivo and in vitro from mono- and multi-PEGylated proteins. Bioconjugate Chem. 17, 341–351. (24) Gitlin, G., Tsarbopoulos, A., Patel, S. T., Sydor, W., Pramanik, B. N., Jacobs, S., Westreich, L., Mittelman, S., and Bausch, J. N. (1996) Isolation and characterization of a monomethioninesulfoxide variant of interferon alpha-2b. Pharm. Res. 13, 762–769.

1020

dx.doi.org/10.1021/bc2001374 |Bioconjugate Chem. 2011, 22, 1017–1020