Native Chemical Ligation of Hydrophobic Peptides in Lipid Bilayer

Apr 24, 2004 - The covalent modification of water-insoluble membrane polypeptides incorporated into lipid bilayers by native chemical ligation is desc...
0 downloads 0 Views 85KB Size
MAY/JUNE 2004 Volume 15, Number 3 © Copyright 2004 by the American Chemical Society

COMMUNICATIONS Native Chemical Ligation of Hydrophobic Peptides in Lipid Bilayer Systems Christie L. Hunter† and Gerd G. Kochendoerfer* Gryphon Therapeutics, 600 Gateway Boulevard, South San Francisco, California 94080. Received February 19, 2004; Revised Manuscript Received March 9, 2004

The covalent modification of water-insoluble membrane polypeptides incorporated into lipid bilayers by native chemical ligation is described. The key feature of this strategy is the use of cubic lipidic phase (CLP) matrixes as reaction media. The CLP-matrix consists of a lipid bilayer into which hydrophobic polypeptides and folded membrane proteins can be inserted and two unbounded aqueous channels that give the aqueous phase access to both sides of an infinite lipid bilayer and thus ensure that modification of solvent-exposed sites is independent of the topology of membrane incorporation. The enzymatic removal of an N-terminal proteolytic cleavage sequence from the membrane polypeptide exposes an N-terminal cysteine residue. Subsequently, a C-terminal thioester peptide is joined to the N-terminus of the polypeptide by a native chemical ligation reaction. By use of this approach, incorporation of a variety of molecular tools, such as spectroscopic probes, unnatural amino acids, and molecular markers into membrane proteins that cannot be easily solubilized in detergent or denaturant solutions, may be achieved.

Chemical synthesis and modification of membrane proteins (1-3) employing chemoselective ligation strategies (4-7) such as native chemical ligation (NCL)1 (5) is a powerful strategy to introduce noncoded amino acids and reporter groups into members of this important class * To whom correspondence should be addressed. Tel: (650) 360 1418; Fax: (650) 952 3055. E-mail: Gkochendoerfer@ gryphonRX.com. † Current address: Applied Biosystems, 850 Lincoln Centre Dr., Foster City, CA 94404. 1 Abbreviations: bR bacteriorhodopsin; CLP, cubic lipidic phase; ESI-MS, electrospray ionization mass spectrometry; FTIR, Fourier transfer infrared; MALDI-TOF, matrix-assisted laser desorption ionization time-of-flight spectrometry; NCL, native chemical ligation; RP-HPLC, reversed-phase highperformance liquid chromatography.

of proteins. Semisynthetic approaches such as expressed protein ligation apply NCL (5) to the linkage of polypeptides in which one or both reaction partners have been produced by recombinant means (8-11). A truncated version of the ion channel protein Kcsa of Streptomyces lividans has been synthesized and subsequently folded by the addition of a synthetic peptide to the C-terminus of a recombinantly produced thioester peptide (12). Typically, such reactions are performed in mixed aqueous-organic solvents, or in the presence of detergents (1-3,12). By contrast, it would be desirable to modify synthetic or recombinant hydrophobic polypeptides in their native lipid environment and thus take advantage of the lipid as the reaction medium or “solvent”. This would be applicable to peptide and bioconjugate chemistry, both for the enzymatic modification of polypeptides

10.1021/bc049959s CCC: $27.50 © 2004 American Chemical Society Published on Web 04/24/2004

438 Bioconjugate Chem., Vol. 15, No. 3, 2004

Figure 1. Schematic of the experimental design demonstrating native chemical ligation in cubic lipidic phases. After incorporation of a peptide corresponding to the first membrane-spanning helix of bacteriorhodopsin (bR) into the lipid bilayer, a peptide containing the protease recognition site was removed by Factor Xa cleavage (i). A peptide tetramer corresponding to the N-terminus of bR was then ligated to the transmembrane peptide to obtain the full bacteriorhodopsin N-terminus (ii). The extent of cleavage and the progress of the ligation inside the CLP was followed by MALDI-TOF mass spectrometry of the respective peptides after extraction with β-D-octylglucopyranoside and affinity capture on streptavidin beads through an N-imino-biotin label (iii). The inset shows a schematic of a cubic lipidic phase (reprinted from plate 9, ref 14, with kind permission from Elsevier).

and for native chemical ligation reactions. Here, we describe a strategy for the introduction of wild-type or labeled peptides to the N-terminus of a synthetic membrane polypeptide in a lipid bilayer environment. As outlined in Figure 1, the enzymatic removal of a proteolytic cleavage sequence from the N-terminus (10, 11) of a membrane polypeptide incorporated into a lipid bilayer generates a free N-terminal cysteine residue. Subsequently, a C-terminal thioester peptide is added to the N-terminus of the protein by a NCL reaction, yielding the desired construct. If desired, ligated polypeptides can be isolated from the lipidic phase after completion of the reaction. A key feature of the approach is the cleavage and ligation reaction of a membrane polypeptide reconstituted into a cubic lipidic phase (CLP). The CLP-matrix consists of a lipid bilayer into which hydrophobic polypeptides and folded membrane proteins can be inserted and two aqueous channels that are both topologically unbounded, giving the aqueous phase access to both sides of an infinite lipid bilayer (13, 14). This feature ensures that all modifications are independent of the topology of membrane incorporation. By contrast, if a more traditional lipid system such as a standard liposome preparation were used for the reaction, the population of peptides that are incorporated with the N-terminus protruding to the inside of the vesicle would not be able to react with extraneously added reactants and vice versa. Since CLPs do not exhibit “inside” or “outside” regions, both the N-terminus and C-terminus of a polypeptide will be amenable for reaction with water-soluble reactants. Since lipid and water diffusion rates inside the CLPmatrix are comparable or slightly slower than those in lamellar lipid phases and bulk water, respectively, sufficient reaction and modification rates can be achieved as also demonstrated in this study (14-16). Incorporation

Hunter and Kochendoerfer

into a CLP also may overcome the difficulty in refolding membrane proteins after ligation reactions under denaturing conditions (1-3, 12, 17), since the reaction can be performed on a membrane polypeptide in the folded state under mild conditions. Finally, a possible shielding effect of detergent molecules that has been observed previously to prevent the ligation reactions of hydrophobic peptides (our own unpublished observation) can be avoided. To demonstrate this approach, we performed an NCL reaction between a synthetic, membrane-incorporated polypeptide (named bRhel)2 derived from the first transmembrane helix of bacteriorhodopsin (bR) of Halobacterium salinarium, and a four-amino acid residue oligopeptide corresponding to the N-terminus of bR (Figure 1). Dissolution of this peptide into typical denaturant buffers was not successful, and ligation thus could not be performed using traditional chemoselective reaction conditions (4-7). By contrast, incorporation into the CLP overcame this solubility problem. To facilitate ligation, Thr5 of bR was replaced with Cys5 to serve as ligation site. This Cys residue was capped with the N-terminal Factor Xa protease recognition site, IEGR (10, 11), to avoid the presence of a free N-terminal cysteine residue during the bilayer reconstitution process and to extend applicability of this approach to recombinant membrane proteins. Alternatively, intein-based expression approaches could be employed to generate proteins with N-terminal cysteines (18). To form the CLP, the bRhel peptide in powder form was mixed with 1-monooleoyl-rac-glycerol (C18:1,[cis]-9) (MO) and 200 mM Tris-HCl buffer (pH 8), forming a cloudy suspension. The suspension was spun in a tabletop centrifuge for 3 h (see Supporting Information for more details). The formation of CLPs was indicated by the formation of optically clear, gellike phases, suggesting that the water-insoluble peptide was incorporated into the lipid phase. The phases remained optically clear at room temperature for at least two months, suggesting high thermodynamic stability. CD and polarized FTIR spectroscopy and proteolysis protection experiments on a bRhel analogue lead to the prediction that this peptide inserts into lipid bilayers in an R-helical structure (19). bRhel proved to be insoluble in aqueous buffers, even in the presence of strong denaturants such as 6 M guanidinium chloride, leading to turbid solutions of aggregated peptide. By contrast, no visible aggregation of hydrophobic peptide was observed after incorporation of bRhel peptide into a CLP at a lipid:peptide ratio of 100:1 and prolonged centrifugation of the CLPs. Taken together, these observations suggest that bRhel is associated with the cubic lipidic phase bilayer, possibly in a helical conformation. Rapid monitoring of the progress of the ligation reaction within the CLP was essential for method development; therefore, an assay was developed that employs affinity capture and subsequent MALDI-TOF mass spectrometry to qualitatively follow the extent of ligation (Figure 2). An N-imino-biotin label was introduced at position Lys41 to facilitate purification and to allow for subsequent affinity capture. MALDI-TOF analysis was performed directly from the affinity beads in order to monitor the ligation reaction immediately without any chromatographic purification step and to avoid errors due to differential ease of release from the affinity beads of 2

The sequence of bRhel is GKGYI EGR

5CG

7RPEWI

12WLALG 17TALMG 22LGTLY 27FLVKG 32MGVSD 37PDAKK.

The residues are numbered according to their position in the sequence of wild-type bacteriorhodopsin.

Communications

Bioconjugate Chem., Vol. 15, No. 3, 2004 439

Figure 3. RP-HPLC chromatogram (detected at 280 nm) after dissolution of the CLP in 60% 2-propanol, 30% acetonitrile, and 10% water containing 0.1% trifluoroacetic acid. The major peak at 9 min correlates with the desired reaction product (observed MW ) 4663 ( 1 Da (theoretical MW ) 4663, average isotope composition). Figure 2. MALDI-TOF analysis of the progress of native chemical ligation in cubic lipidic phases. (A) MALDI-TOF spectra of bRhel prior to protease cleavage. (B) MALDI-TOF spectra of this peptide after protease cleavage. (C-E) MALDITOF spectra of this peptide after ligation with peptide EAQL after 25 min (C), 75 min (D), and overnight (E). The peaks of varying intensity marked with an asterisk are due to impurities associated with the purification and analysis process. Spectra were obtained directly from the affinity beads plated on MALDI slides.

the various constructs. The uneven surface of the matrix spot due to the presence of the beads on the plate is the cause of the low resolution and mass accuracy observed in the mass spectra shown. Spectrum A presents a MALDI-TOF spectrum of bRhel after incorporation into a CLP, followed by dissolution of the CLP in a buffer containing 20 mM Bis-Tris Propane buffer (1,3-bis[tris(hydroxymethyl)methylamino]propane) (pH ) 9.5), 27% β-D-octylglucopyranoside, and 20% β-mercaptoethanol, and affinity capture onto streptavidin-coated beads. The reducing agent was added to keep all peptide components and ligation products in the reduced, disulfide-free state during analysis and workup. The spectrum displays one major peak at m/z ) 5081, corresponding to the mass of uncleaved bRhel. Smaller peaks (labeled with an asterisk) represent impurities from the lipid matrix, extraction process, and/or the affinity beads and were also present in the mass spectra in the absence of protein (data not shown). Whereas these impurities vary in intensity between individual timepoints due to variations in the number of beads spotted on the plate, they do not interfere with the peaks of most interest and therefore the monitoring of the progress of the reaction. Spectrum B presents a mass spectrum of the same peptide after cleavage with Factor Xa inside the CLP overnight. The disappearance of the major peak at m/z ) 5081 concomitant with the appearance of a major peak at m/z ) 4222 is consistent with efficient removal of the Cys-capping sequence GKGYIEGR by the Factor Xa protease. Spectra C, D, and E show the appearance of ligation product after addition of a 2-fold excess of N-terminal peptide EAQL thioester with 0.5% thiophenol to the CLP after 25 and 75 min and overnight incubation, respectively. The gradual disappearance of the peak at m/z ) 4222 Da with a concomitant increase of a peak at m/z ) 4664 Da demonstrates the progress of the formation of a

native amide bond between the bR transmembrane peptide and the bR N-terminus. After the reaction was complete, the reaction mixture was further analyzed and purified by RP-HPLC and subsequent electrospray mass spectrometry. For this analysis, the CLP was dissolved in 60% 2-propanol, 30% acetonitrile, and 10% water containing 0.1% trifluoroacetic acid. The analytical RP-HPLC chromatogram of an aliquot of the resultant solution is presented in Figure 3. The chromatogram exhibits a dominant peak at ∼ 9 min retention time. ESI-MS analysis of this peak indicates this component has the same mass as the desired reaction product (4663 ( 1 Da, see Figure 3 inset) and that it can be isolated with high purity. Reactions were typically performed at a 600 µg scale starting material (see Supporting Information), and the overall yield of pure product purified by RP-HPLC was typically ∼33% (200 µg). Preparation of CLPs is scalable and is expected to be amenable to multi-milligram scale modification of hydrophobic polypeptides. The synthetic strategy described in this communication has several applications in the synthesis and modification of hydrophobic polypeptides. The limited solubility of hydrophobic membrane polypeptides in aqueous solvents has been a challenge to their modification (1-3, 12). Application of lipid bilayer systems, in particular CLPs, may make chemoselective ligation techniques more applicable to this important class of peptides. Our approach may allow researchers to modularly assemble such membrane proteins from separately synthesized (or recombinantly expressed) subunits. In particular, multiplemembrane spanning proteins can be assembled from individual membrane-spanning domains without possible worries of incorporation of the peptides in incompatible relative orientation due to the unique topological features of the CLP system. Incorporation of a variety of molecular tools, such as spectroscopic probes, unnatural amino acids, and molecular markers into membrane proteins at yields that significantly exceed those accessible with currently available techniques, may thus be achieved. NOTE ADDED AFTER PRINT PUBLICATION

In the manuscript title, the word “Hydrophobic” was misspelled in the version of this paper published 4/24/2004 (ASAP) and in the May/June 2004 issue

440 Bioconjugate Chem., Vol. 15, No. 3, 2004

(Vol. 15, No. 3, pp 437-440). The correct electronic version was published on the Web 6/4/2004, and a Correction appears in the July/August 2004 issue (Vol. 15, No. 4). ACKNOWLEDGMENT

Hunter and Kochendoerfer (8) Muir, T. W. (2003) Semisynthesis of proteins by expressed protein ligation. Annu. Rev. Biochem. 72, 249-289. (9) 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.

We thank Stephen B.H. Kent (University of Chicago) for advice and encouragement.

(10) Xu, R., Ayers, B., Cowburn, D., and Muir, T. W. (1999) Chemical ligation of folded recombinant proteins: segmental isotopic labeling of domains for NMR studies. Proc. Natl. Acad. Sci. U.S.A. 96, 388-393.

Supporting Information Available: Detailed descriptions of experimental procedures. This material is available free of charge via the Internet at http://pubs.acs.org.

(11) Erlanson, D. A., Chytil, M., and Verdine, G. L. (1996) The leucine zipper domain controls the orientation of AP-1 in the NFAT.AP-1.DNA complex. Chem. Biol. 3, 981-991.

LITERATURE CITED (1) Kochendoerfer, G. G., Salom, D., Lear, J. D., Wilk-Orescan, R., Kent, S. B. H., and DeGrado, W. F. (1999) Total Chemical Synthesis of the Integral Membrane Protein Influenza A Virus M2: Role of Its C-Terminal Domain in Tetramer Assembly. Biochemistry 38, 11905-11913. (2) Clayton, D., Shapovalov, G., Maurer, J. H., Dougherty, D. A., Lester, H. A., and Kochendoerfer, G. G. (2004) Total chemical synthesis and electrophysiological characterization of mechanosensitive channels from Escherichia coli and Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. 101, 4764-4769. (3) Kochendoerfer, G., Jones, D., Lee, S., Oblatt-Montal, M., Opella, S., and Montal, M. (2004) Functional characterization and NMR spectroscopy on full-length Vpu from HIV-1 prepared by total chemical synthesis. J. Am. Chem. Soc. 137, 2439-2446. (4) Schno¨lzer, M., and Kent, S. B. H. (1992) Constructing Proteins by Dovetailing Unprotected Synthetic Peptides: Backbone-Engineered HIV Protease. Science 256, 221-225. (5) 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. (6) Dawson, P. E., and Kent, S. B. H. (2000) Synthesis of native proteins by chemical ligation. Annu. Rev. Biochem 69, 923962. (7) Kochendoerfer, G. G. (2001) Chemical protein synthesis methods in drug discovery. Curr. Opin. Drug. Discovery Dev. 4, 205-214.

(12) Valiyaveetil, F. I., MacKinnon, R., and Muir, T. W. (2002) Semisynthesis and folding of the potassium channel KcsA. J. Am. Chem. Soc. 124, 9113-9120. (13) Luzzatti, V., Gulik-Krzywicki, T., and Tardieu, A. (1968) Polymorphism of lecithins. Nature 218, 1031-1034. (14) Lindblom, G., and Rilfors, L. (1989) Cubic phases and isotropic structures formed by membrane lipids- possible biological relevance. Biochim. Biophys. Acta 988, 221-256. (15) Rowinski, P., Korytkowska, A., and Bilewicz, R. (2003) Diffusion of hydrophilic probes in bicontinuous lipidic cubic phase. Chem. Phys. Lip. 124, 147-156. (16) Portmann, M., Landau, E. M., and Luisi, P. L. (1991) Spectroscopic and Rheological Studies of Enzymes in Rigid Lipidic Matrixes: The Case of R-Chymotrypsin in a Lysolecithin/Water cubic Phase. J. Phys. Chem. 95, 8437-8440. (17) Valiyaveetil, F., Zhou, Y., and MacKinnon, R. (2002) Lipids in the structure, folding, and function of the KcsA K+ channel. Biochemistry 41, 10771-10777. (18) Perler, F. B., Xu, M. Q., and Paulus, H. (1997) Protein splicing and autoproteolysis mechanisms. Curr. Opin. Chem. Biol. 1, 292-299. (19) Hunt, J. F., Bousche, O., Meyers, K. M., Rothschild, K. J., and Engelman, D. M. (1991) Biophysical studies of the integral membrane protein folding pathway. Biophys. J. 59, 400a.

BC049959S