- Email: [email protected]
Small-molecule catalysts of oxidative protein folding
Available online at www.sciencedirect.com
Small-molecule catalysts of oxidative protein folding Watson J Lees Oxidative protein folding occurs both in vivo and in vitro and involves the formation and rearrangement of protein disulfide bonds (–S–S– bonds). In vivo these reactions are catalyzed by enzymes, including the eukaryotic enzyme protein disulfide isomerase (PDI). Using the physical properties of PDI as a guide, several small-molecule catalysts of oxidative protein folding have been designed, synthesized, and tested. These small molecules can improve the folding rate of the model substrate ribonuclease A by a factor of over 10 and improve the yield by up to a factor of 3 over traditional conditions. The molecules have also been demonstrated to significantly improve the in vivo folding of proteins as well. Address Department of Chemistry and Biochemistry, Florida International University, 11200 SW 8th Street, Miami, FL 33199, USA Corresponding author: Lees, Watson J ([email protected])
Current Opinion in Chemical Biology 2008, 12:740–745 This review comes from a themed issue on Model Systems Edited by Helma Wennemers and Ronald T. Raines Available online 25th September 2008 1367-5931/$ – see front matter # 2008 Elsevier Ltd. All rights reserved. DOI 10.1016/j.cbpa.2008.08.032
In both cases the formation of disulfide bonds is accelerated by the presence of chaperones such as protein disulfide isomerase (PDI) or a series of disulfide bond formation proteins, also known as Dsb proteins. About 20% of all human proteins are predicted to contain disulfide bonds [1]. Inhibition of PDI is proposed to be a factor in a number of diseases [2]. The in vitro folding of proteins is important in both the research laboratory and commercially [3]. The production of proteins, especially disulfide-containing proteins, in the most commonly used organism E. coli can result in the formation of protein aggregates/precipitates called inclusion bodies instead of folded native protein. The formation of inclusion bodies can be advantageous as it allows for higher expression levels, simplified purification schemes, and resistance to cellular proteases. The disadvantage is that the inactive protein within the inclusion bodies needs to be resolubilized using denaturants such as urea and then folded in vitro to form native protein. For disulfide-containing proteins, in vitro folding corresponds to oxidative protein folding and almost always represents the most challenging step. This review focuses on the recently developed smallmolecule catalysts of oxidative protein folding. These molecules affect the underlying thiol–disulfide interchange reactions that occur during folding, such as the rearrangement of disulfide bonds or the oxidation of protein thiols to protein disulfides. However, unlike ‘true’ catalysts, the small molecules do sometimes provide oxidizing equivalents.
Introduction Design
Disulfide bonds (–S–S– bonds) represent an important and in many cases an essential structural feature of numerous proteins, especially extracellular ones. Disulfide bonds are formed by the oxidation of cysteine residues and provide stability to the protein by minimizing the entropy of the unfolded state as they crosslink one part of a protein with another. The formation of native disulfide bonds in vivo or in vitro occurs through oxidative protein folding. Oxidative protein folding involves conformational folding combined with the oxidation of protein thiols to disulfides, the reduction of protein disulfides to thiols, and the rearrangement of non-native disulfide bonds to native disulfide bonds, ultimately culminating in the production of native protein (Figure 1).
The inspiration for the design of small-molecule catalysts of oxidative protein folding has come from nature, especially from PDI. PDI is a eukaryotic protein, found in high concentrations in the ER, that catalyzes oxidative protein folding in vivo and in vitro [4]. PDI contains two active sites, each of which has a CXXC motif, where X is any amino acid and C is cysteine [5]. The two-thiol containing cysteines form a 14-membered ring upon oxidation to the disulfide. The stability of the disulfide is measured in terms of its reduction potential, E80 . The more stable the disulfide the lower the E80 . For PDI the value is 180 mV [6]. One of the active-site thiols has a low thiol pKa of 6.7, and the corresponding thiolate (deprotonated thiol) displays enhanced reactivity toward disulfide bonds [7,8].
Oxidative protein folding is important both in vivo and in vitro. In vivo, oxidative protein folding mainly occurs inside the endoplasmic reticulum (ER) of eukaryotes or in the periplasmic space of Gram-negative bacteria.
The properties of having two thiol groups, a low thiol pKa, a low E80 , and enhanced reactivity are expected to be the important advantages for protein folding catalysts. Having two thiol groups is proposed to minimize the
Current Opinion in Chemical Biology 2008, 12:740–745
www.sciencedirect.com
Small-molecule catalysts of oxidative protein folding Lees 741
Figure 1
Intramolecular rearrangement of non-native to native protein disulfide bonds and the oxidation of a protein dithiol to a protein disulfide using a smallmolecule disulfide.
possibility of forming long-lived mixed disulfides between the folding catalyst and the protein of interest. If the mixed disulfide persists for too long, the second thiol will intramolecularly remove the folding catalyst [9,10]. This is not possible with monothiols and has
been called the escape mechanism (Figure 2). A low pKa thiol is advantageous, as the most reactive small-molecule nucleophile in solution is the one with a pKa value close to the pH of solution, all else being equal. Usually, aliphatic and cysteine-based thiols have pKa values somewhat
Figure 2
The proposed advantage of having two thiols in each active site of PDI, the escape mechanism. www.sciencedirect.com
Current Opinion in Chemical Biology 2008, 12:740–745
742 Model Systems Figure 3
CXXC corresponds to the active-site sequence of thioredoxin [His37] [31–38], cyclic CXXC corresponds to the active-site sequence of PDI [35–40] with CGHC being the active site as indicated by circles, and the photoactive peptide corresponds to the active-site sequence of thioredoxin reductase [131–141]. The –N N– bond of the photoactive peptide undergoes cis–trans isomerization upon exposure to light. Ox and Red correspond to oxidation and reduction, respectively. Current Opinion in Chemical Biology 2008, 12:740–745
www.sciencedirect.com
Small-molecule catalysts of oxidative protein folding Lees 743
greater than physiological pH. A low E80 value allows PDI to exist as a nearly 50:50 mixture of dithiol and disulfide in the ER and under in vitro conditions that fold many proteins efficiently [11]. PDI in the oxidized disulfide form can oxidize protein thiols to disulfides; in the reduced form PDI can help rearrange non-native disulfide bonds to native disulfide bonds.
Small-molecule catalysts The active sites of PDI and related enzymes have been modeled by small linear, cyclic, or photoactive CXXC peptides that have a variety of reduction potentials and slightly lower than normal thiol pKa values (Figure 3; Table 1) [12,13,14]. The advantage of the cyclic peptides over linear peptides is that they can form more strained disulfide bonds, with consequently higher values of E80 ; the photoactive peptide is cyclized with an azobenzene that changes the conformation of the peptide and thus the reduction potential when it is converted from the cis to the trans form with light. In addition, CXXC peptides have been immobilized on multiple antigen peptide (MAP) resin and successfully used for protein folding. However, the properties listed in Table 1 were not reported [15]. The active site has also been modeled as a CXC peptide, which upon oxidation forms a strained 11-membered disulfide ring [16]. All of these peptides contain two thiol groups; in Table 1 the most efficient catalyst of each series was selected. Small molecules that do not model the active sites of PDI but incorporate many of the highlighted properties of PDI have also been used as protein folding catalysts. ()-trans1,2-Bis(2-mercaptoacetamido)cyclohexane (BMC) is a dithiol with a lower thiol pKa value and higher redox potential than glutathione [10]. Aromatic thiols incorporate the lower thiol pKa of PDI as well as the enhanced reactivity of the corresponding thiolate. Aromatic thiolates are 5–10 times more reactive toward small-molecule disulfides than their aliphatic counterparts with similar thiol pKa values [17]. The thiol pKa value can be tuned by varying the substituents on the aromatic ring from around 5 to 7. Tuning of the thiol pKa value also results in a
change in the reduction potential at pH 7 (E80 ), as the equilibrium is shifted by deprotonation of the thiol. The selenide analog of GSH, GSeH, has a low selenol pKa value and enhanced reactivity in both the diselenide and the selenide forms [18]. The diselenide is very reactive with thiolates and the selenide reacts rapidly with disulfides, even though a diselenide bond is intrinsically more stable than a disulfide bond. Furthermore, thiols have also been attached to microspheres and used successfully to enhance the folding of proteins, compared to nonthiol containing microspheres [19].
Protein folding The traditional model system for protein folding is ribonuclease A (RNase A). Experiments were initiated with RNase A in either the reduced denatured form (rRNase A), which contains eight cysteine thiols, or in the scrambled form (sRNase A), which has a random distribution of its four disulfide bonds (Table 1) [20,21,22]. Neither form possesses appreciable enzymatic activity. Upon the addition of either RNase A form to the folding mixture, the RNase A slowly regains enzymatic activity, usually over hours. A plot of enzymatic activity versus time is then fit to a single exponential, A(1 ekt), with A being the enzymatic activity observed at long times and k being the first-order folding rate constant (Figure 4). The value of A is also called the yield when expressed as a percent of the activity obtained if all the reduced or scrambled RNase A folded only to native RNase A. For comparing reagents, the standard conditions are 1 mM GSSG and 0.2 mM GSH. Most of the reagents have been compared to the standard conditions, though the pH and temperature varied somewhat [23]. Several of the folding catalysts have also been used to fold additional proteins in vitro and in vivo. BMC was shown to increase the in vivo folding yield of a secreted acid phosphatase in yeast threefold upon addition to the growth medium [10]. A more recent report using a fractional factorial screen showed that BMC performed better than both GSH/GSSG and dithiothreitol, DTT, at folding a range of disulfide-containing proteins in terms of
Table 1 Folding of reduced (r) or scrambled (s) RNase Aa Reagent Linear CXXC Cyclic CXXC photo CXXC cis/trans CXC or CGC BMC Aromatic 1 Aromatic 2 GSeH GSH a
Thiol pKa na 7.3, 9.6 na 8.7, 9.8 8.3, 9.9 6.6 5.7 5.3 (GSeH) 8.7
E80 (mV) (disulfide) 190 130 147/201 167 240 220 170 407 250
pH, T (8C), RNase A 7.4, 7.4, 7.4, 7.6, 7.6, 7.0, 7.0, 8.0,
20, 30, 30, 30, 30, 25, 25, 25,
r r r s s s s r
k (reagent/GSH) Twofold Twofold Approx. 1 Onefold to twofold Approx. 1 Sevenfold 12-fold Twofold
Yield (reagent/GSH) Approx. same 95%/83% 101%/77% 60%/20% 90%/40% Approx. same Approx. same Approx. same
Italics represent the best guess from the folding curves provided while na means that the data were not available.
www.sciencedirect.com
Current Opinion in Chemical Biology 2008, 12:740–745
744 Model Systems
Figure 4
Folding of scrambled RNase A (0.025 mM) at 25 8C and pH 7.7 in the presence of 1.0 mM GSH and 0.2 mM GSSG (standard conditions). The enzymatic activity was followed at discrete time points and fit to A(1 ekt), A and k values correspond to 91% yield and 0.008 min1, respectively.
yield [24]. Even at low concentrations, BMC increased the in vitro folding yield and rate. The in vitro folding rate of rRNase A by DTTox is enhanced twofold by the addition of 50 micromolar BMC [25]. At 10 micromolar BMC, the in vitro folding rate of proinsulin in the presence of GSH and GSSG increased only slightly but the yield of active protein went up from 20 to 40% [26]. Higher concentrations of BMC decreased the yield. It was concluded that BMC works at stages that affect aggregation but not folding rate. BMC could also be added to the growth medium of E. coli (2–50 micromolar) to increase the production of proinsulin by 60%. Again, higher concentrations were disadvantageous. Aromatic disulfides have been used previously for the rapid oxidation of thiols to disulfides, but until recently they have not been used for oxidative protein folding [27]. A series of aromatic thiols with a range of thiol pKa values were synthesized and tested for their ability to fold sRNase A. The effects of thiol pKa, thiol concentration, and solution pH on the folding rates were determined, and a general equation was obtained to predict folding rates [28]. Ultimately, a lower thiol pKa led to a faster folding rate under optimal condition, but higher thiol concentrations were required to reach optimal conditions. A thiol pKa approximately 1 unit less than the pH was viewed as optimal. Aromatic thiols were also demonstrated to enhance the folding rate of scrambled RNase A by threefold in the presence of PDI, though factors other than thiol pKa appeared to be important [29]. In addition, aromatic thiols were shown to improve the yield from 70 to 85% and the folding rate for lysozyme up to 11fold compared to glutathione [30]. Furthermore, an aromatic thiol, thiosalicylic acid, folded a single-chain antiCurrent Opinion in Chemical Biology 2008, 12:740–745
body fragment twice as fast as glutathione and increased the yield from 11 to 19% at pH 7 [31]. Immobilized Ellman’s reagent, an aromatic disulfide also known as CLEAR-OX, improved the folding yield of conotoxins containing two disulfide bonds by up to 10-fold or 20-fold, especially at high peptide concentrations, and improved the yield of conotoxins containing three disulfide bonds from 21 to 32% over those obtained in solution [32,33]. The effect of GSeSeG on the folding of bovine pancreatic trypsin inhibitor (BPTI) was more dramatic than that observed with RNase A [18]. At short times the amount of native protein produced was similar for both GSSG and GSeSeG, but at longer times GSeSeG provided significantly more native protein. Presumably, with GSeSeG the last native disulfide bond was formed via direct oxidation, as air was reported to enhance the reaction rate, whereas with GSSG it was formed by intramolecular rearrangement of disulfide bonds within the protein followed by oxidation [34]. Recently, substoichiometric amounts of GSeSeG have been shown to enhance RNase A folding relative to GSSG by catalyzing the oxidation folding of proteins with oxygen [35].
Conclusion In general, the small-molecule dithiols presented above will improve the yield of active protein. Aromatic thiols and selenols will enhance the protein folding rates. However, neither the aromatic thiols nor the selenols have been tested under conditions where the yield of active protein is low using standard conditions. The expectation is that small-molecule catalysts of oxidative protein folding will only continue to improve as more and more of the properties of PDI are incorporated into a single small molecule. These new catalysts will enhance the production of protein both in vivo and in vitro. As the link between oxidative protein folding and diseases is clarified, their in vivo applications will expand.
Acknowledgements The author would like to thank Drs David Chatfield, Fenfei Leng, and Alex Mebel for their helpful suggestions. Financial support from the National Science Foundation (CHE-0342167 to WJL) is gratefully acknowledged.
References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as: of special interest of outstanding interest 1.
Martelli PL, Fariselli P, Casadio R: Prediction of disulfide-bonded cysteines in proteomes with a hidden neural network. Proteomics 2004, 4:1665-1671.
2.
Nakamura T, Lipton SA: Emerging roles of S-nitrosylation in protein misfolding and neurodegenerative diseases. Antioxid Redox Signal 2008, 10:87-101.
3.
Sahdev S, Khattar SK, Saini KS: Production of active eukaryotic proteins through bacterial expression systems: a review of the existing biotechnology strategies. Mol Cell Biochem 2008, 307:249-264. www.sciencedirect.com
Small-molecule catalysts of oxidative protein folding Lees 745
4.
Kersteen EA, Raines RT: Catalysis of protein folding by protein disulfide isomerase and small-molecule mimics. Antioxid Redox Signal 2003, 5:413-424. A nice review of the properties and function of PDI.
19. Woycechowsky KJ, Hook BA, Raines RT: Catalysis of protein folding by an immobilized small-molecule dithiol. Biotechnol Progr 2003, 19:1307-1314.
5.
20. Crook EM, Mathias AP, Rabin BR: Spectrophotometric assay of bovine pancreatic ribonuclease by the use of cytidine 20 :30 phosphate. Biochem J 1960, 74:234-238.
6.
21. Konishi Y, Scheraga HA: Regeneration of ribonuclease A from the reduced protein. 1. Conformational analysis of the intermediates by measurements of enzymic activity, optical density, and optical rotation. Biochemistry 1980, 19:1308-1316.
Tian G, Xiang S, Noiva R, Lennarz WJ, Schindelin H: The crystal structure of yeast protein disulfide isomerase suggests cooperativity between its active sites. Cell (Cambridge, MA, USA) 2006, 124:61-73. The crystal structure of PDI.
7.
8.
9.
Lundstroem J, Holmgren A: Determination of the reduction– oxidation potential of the thioredoxin-like domains of protein disulfide-isomerase from the equilibrium with glutathione and thioredoxin. Biochemistry 1993, 32:6649-6655. Hawkins HC, Freedman RB: The reactivities and ionization properties of the active-site dithiol groups of mammalian protein disulfide-isomerase. Biochem J 1991, 275:335-339. Darby N, Creighton TE: Characterization of the active site cysteine residues of the thioredoxin-like domains of protein disulfide isomerase. Biochemistry 1995, 34:16770-16780. Walker KW, Gilbert HF: Scanning and escape during proteindisulfide isomerase-assisted protein folding. J Biol Chem 1997, 272:8845-8848.
10. Woycechowsky KJ, Wittrup KD, Raines RT: A small-molecule catalyst of protein folding in vitro and in vivo. Chem Biol 1999, 6:871-879. This paper describes the first use of a nonpeptide small-molecule dithiol for the folding of RNase A and demonstrates the successful in vivo application of a small-molecule catalyst of oxidative protein folding. 11. Appenzeller-Herzog C, Ellgaard L: In vivo reduction–oxidation state of protein disulfide isomerase: the two active sites independently occur in the reduced and oxidized forms. Antioxid Redox Signal 2008, 10:55-64. 12. Moroder L, Besse D, Musiol H-J, Rudolph-Boehner S, Sideler F: Oxidative folding of cystine-rich peptides versus regioselective cysteine pairing strategies. Biopolymers 1996, 40:207-234. 13. Cabrele C, Fiori S, Pegoraro S, Moroder L: Redox-active cyclic bis(cysteinyl)peptides as catalysts for in vitro oxidative protein folding. Chem Biol 2002, 9:731-740. This paper depicts the use of cyclic CXXC peptides that better mimic the properties of PDI than the original linear CXXC peptides. 14. Cabrele C, Cattani-Scholz A, Renner C, Behrendt R, Oesterhelt D, Moroder L: Photomodulation of the redox and folding adjuvant properties of bis(cysteinyl) peptides. Eur J Org Chem 2002, 2002:2144-2150. 15. Ookura T, Kainuma K, Kim H-j, Otaka A, Fujii N, Kawamura Y: Active site peptides with CXXC motif on MAP-resin can mimic protein disulfide isomerase activity. Biochem Biophys Res Commun 1995, 213:746-751. 16. Woycechowsky KJ, Raines RT: The CXC motif: a functional mimic of protein disulfide isomerase. Biochemistry 2003, 42:5387-5394. 17. DeCollo TV, Lees WJ: Effects of aromatic thiols on thiol– disulfide interchange reactions that occur during protein folding. J Org Chem 2001, 66:4244-4249. 18. Beld J, Woycechowsky KJ, Hilvert D: Selenoglutathione: efficient oxidative protein folding by a diselenide. Biochemistry 2007, 46:5382-5390. This account portrays the first use of the more reactive diselenides to fold proteins.
www.sciencedirect.com
22. Gough JD, Williams RH Jr, Donofrio AE, Lees WJ: Folding disulfide-containing proteins faster with an aromatic thiol. J Am Chem Soc 2002, 124:3885-3892. This paper describes the first use of an aromatic thiol for the folding of disulfide-containing proteins. 23. Lyles MM, Gilbert HF: Catalysis of the oxidative folding of ribonuclease A by protein disulfide isomerase: dependence of the rate on the composition of the redox buffer. Biochemistry 1991, 30:613-619. 24. Willis MS, Hogan JK, Prabhakar P, Liu X, Tsai K, Wei Y, Fox T: Investigation of protein refolding using a fractional factorial screen: a study of reagent effects and interactions. Protein Sci 2005, 14:1818-1826. 25. Fink M, Nieves P, Chang S, Narayan M: Non-redox-active smallmolecules can accelerate oxidative protein folding by novel mechanisms. Biophys Chem 2008, 132:104-109. 26. Winter J, Lilie H, Rudolph R: Recombinant expression and in vitro folding of proinsulin are stimulated by the synthetic dithiol Vectrase-P. FEMS Microbiol Lett 2002, 213:225-230. 27. Annis I, Chen L, Barany G: Novel solid-phase reagents for facile formation of intramolecular disulfide bridges in peptides under mild conditions. J Am Chem Soc 1998, 120:7226-7238. 28. Gough JD, Lees WJ: Effects of redox buffer properties on the folding of a disulfide-containing protein: dependence upon pH, thiol pKa, and thiol concentration. J Biotechnol 2005, 115:279-290. This is a more detailed account of the use of aromatic thiols to fold proteins and provides some understanding of the important factors. 29. Gough JD, Lees WJ: Increased catalytic activity of protein disulfide isomerase using aromatic thiol based redox buffers. Bioorg Med Chem Lett 2005, 15:777-781. 30. Gurbhele-Tupkar MC, Perez LR, Silva Y, Lees WJ: Rate enhancement of the oxidative folding of lysozyme by the use of aromatic thiol containing redox buffers. Bioorg Med Chem 2008, 16:2579-2590. 31. Patil G, Rudolph R, Lange C: In vitro-refolding of a single-chain Fv fragment in the presence of heteroaromatic thiols. J Biotechnol 2008, 134:218-221. 32. Darlak K, Long DW, Czerwinski A, Darlak M, Valenzuela F, Spatola AF, Barany G: Facile preparation of disulfide-bridged peptides using the polymer-supported oxidant CLEAR-OX. J Pept Res 2004, 63:303-312. 33. Green BR, Bulaj G: Oxidative folding of conotoxins in immobilized systems. Protein Pept Lett 2006, 13:67-70. 34. Kibria FM, Lees WJ: Balancing conformational and oxidative kinetic traps during the folding of bovine pancreatic trypsin inhibitor (BPTI) with glutathione and glutathione disulfide. J Am Chem Soc 2008, 130:796-797. 35. Beld J, Woycechowsky KJ, Hilvert D: Catalysis of oxidative protein folding by small-molecule diselenides. Biochemistry 2008, 47:6985-6987.
Current Opinion in Chemical Biology 2008, 12:740–745
Small-molecule catalysts of oxidative protein folding Watson J Lees Oxidative protein folding occurs both in vivo and in vitro and involves the formation and rearrangement of protein disulfide bonds (–S–S– bonds). In vivo these reactions are catalyzed by enzymes, including the eukaryotic enzyme protein disulfide isomerase (PDI). Using the physical properties of PDI as a guide, several small-molecule catalysts of oxidative protein folding have been designed, synthesized, and tested. These small molecules can improve the folding rate of the model substrate ribonuclease A by a factor of over 10 and improve the yield by up to a factor of 3 over traditional conditions. The molecules have also been demonstrated to significantly improve the in vivo folding of proteins as well. Address Department of Chemistry and Biochemistry, Florida International University, 11200 SW 8th Street, Miami, FL 33199, USA Corresponding author: Lees, Watson J ([email protected])
Current Opinion in Chemical Biology 2008, 12:740–745 This review comes from a themed issue on Model Systems Edited by Helma Wennemers and Ronald T. Raines Available online 25th September 2008 1367-5931/$ – see front matter # 2008 Elsevier Ltd. All rights reserved. DOI 10.1016/j.cbpa.2008.08.032
In both cases the formation of disulfide bonds is accelerated by the presence of chaperones such as protein disulfide isomerase (PDI) or a series of disulfide bond formation proteins, also known as Dsb proteins. About 20% of all human proteins are predicted to contain disulfide bonds [1]. Inhibition of PDI is proposed to be a factor in a number of diseases [2]. The in vitro folding of proteins is important in both the research laboratory and commercially [3]. The production of proteins, especially disulfide-containing proteins, in the most commonly used organism E. coli can result in the formation of protein aggregates/precipitates called inclusion bodies instead of folded native protein. The formation of inclusion bodies can be advantageous as it allows for higher expression levels, simplified purification schemes, and resistance to cellular proteases. The disadvantage is that the inactive protein within the inclusion bodies needs to be resolubilized using denaturants such as urea and then folded in vitro to form native protein. For disulfide-containing proteins, in vitro folding corresponds to oxidative protein folding and almost always represents the most challenging step. This review focuses on the recently developed smallmolecule catalysts of oxidative protein folding. These molecules affect the underlying thiol–disulfide interchange reactions that occur during folding, such as the rearrangement of disulfide bonds or the oxidation of protein thiols to protein disulfides. However, unlike ‘true’ catalysts, the small molecules do sometimes provide oxidizing equivalents.
Introduction Design
Disulfide bonds (–S–S– bonds) represent an important and in many cases an essential structural feature of numerous proteins, especially extracellular ones. Disulfide bonds are formed by the oxidation of cysteine residues and provide stability to the protein by minimizing the entropy of the unfolded state as they crosslink one part of a protein with another. The formation of native disulfide bonds in vivo or in vitro occurs through oxidative protein folding. Oxidative protein folding involves conformational folding combined with the oxidation of protein thiols to disulfides, the reduction of protein disulfides to thiols, and the rearrangement of non-native disulfide bonds to native disulfide bonds, ultimately culminating in the production of native protein (Figure 1).
The inspiration for the design of small-molecule catalysts of oxidative protein folding has come from nature, especially from PDI. PDI is a eukaryotic protein, found in high concentrations in the ER, that catalyzes oxidative protein folding in vivo and in vitro [4]. PDI contains two active sites, each of which has a CXXC motif, where X is any amino acid and C is cysteine [5]. The two-thiol containing cysteines form a 14-membered ring upon oxidation to the disulfide. The stability of the disulfide is measured in terms of its reduction potential, E80 . The more stable the disulfide the lower the E80 . For PDI the value is 180 mV [6]. One of the active-site thiols has a low thiol pKa of 6.7, and the corresponding thiolate (deprotonated thiol) displays enhanced reactivity toward disulfide bonds [7,8].
Oxidative protein folding is important both in vivo and in vitro. In vivo, oxidative protein folding mainly occurs inside the endoplasmic reticulum (ER) of eukaryotes or in the periplasmic space of Gram-negative bacteria.
The properties of having two thiol groups, a low thiol pKa, a low E80 , and enhanced reactivity are expected to be the important advantages for protein folding catalysts. Having two thiol groups is proposed to minimize the
Current Opinion in Chemical Biology 2008, 12:740–745
www.sciencedirect.com
Small-molecule catalysts of oxidative protein folding Lees 741
Figure 1
Intramolecular rearrangement of non-native to native protein disulfide bonds and the oxidation of a protein dithiol to a protein disulfide using a smallmolecule disulfide.
possibility of forming long-lived mixed disulfides between the folding catalyst and the protein of interest. If the mixed disulfide persists for too long, the second thiol will intramolecularly remove the folding catalyst [9,10]. This is not possible with monothiols and has
been called the escape mechanism (Figure 2). A low pKa thiol is advantageous, as the most reactive small-molecule nucleophile in solution is the one with a pKa value close to the pH of solution, all else being equal. Usually, aliphatic and cysteine-based thiols have pKa values somewhat
Figure 2
The proposed advantage of having two thiols in each active site of PDI, the escape mechanism. www.sciencedirect.com
Current Opinion in Chemical Biology 2008, 12:740–745
742 Model Systems Figure 3
CXXC corresponds to the active-site sequence of thioredoxin [His37] [31–38], cyclic CXXC corresponds to the active-site sequence of PDI [35–40] with CGHC being the active site as indicated by circles, and the photoactive peptide corresponds to the active-site sequence of thioredoxin reductase [131–141]. The –N N– bond of the photoactive peptide undergoes cis–trans isomerization upon exposure to light. Ox and Red correspond to oxidation and reduction, respectively. Current Opinion in Chemical Biology 2008, 12:740–745
www.sciencedirect.com
Small-molecule catalysts of oxidative protein folding Lees 743
greater than physiological pH. A low E80 value allows PDI to exist as a nearly 50:50 mixture of dithiol and disulfide in the ER and under in vitro conditions that fold many proteins efficiently [11]. PDI in the oxidized disulfide form can oxidize protein thiols to disulfides; in the reduced form PDI can help rearrange non-native disulfide bonds to native disulfide bonds.
Small-molecule catalysts The active sites of PDI and related enzymes have been modeled by small linear, cyclic, or photoactive CXXC peptides that have a variety of reduction potentials and slightly lower than normal thiol pKa values (Figure 3; Table 1) [12,13,14]. The advantage of the cyclic peptides over linear peptides is that they can form more strained disulfide bonds, with consequently higher values of E80 ; the photoactive peptide is cyclized with an azobenzene that changes the conformation of the peptide and thus the reduction potential when it is converted from the cis to the trans form with light. In addition, CXXC peptides have been immobilized on multiple antigen peptide (MAP) resin and successfully used for protein folding. However, the properties listed in Table 1 were not reported [15]. The active site has also been modeled as a CXC peptide, which upon oxidation forms a strained 11-membered disulfide ring [16]. All of these peptides contain two thiol groups; in Table 1 the most efficient catalyst of each series was selected. Small molecules that do not model the active sites of PDI but incorporate many of the highlighted properties of PDI have also been used as protein folding catalysts. ()-trans1,2-Bis(2-mercaptoacetamido)cyclohexane (BMC) is a dithiol with a lower thiol pKa value and higher redox potential than glutathione [10]. Aromatic thiols incorporate the lower thiol pKa of PDI as well as the enhanced reactivity of the corresponding thiolate. Aromatic thiolates are 5–10 times more reactive toward small-molecule disulfides than their aliphatic counterparts with similar thiol pKa values [17]. The thiol pKa value can be tuned by varying the substituents on the aromatic ring from around 5 to 7. Tuning of the thiol pKa value also results in a
change in the reduction potential at pH 7 (E80 ), as the equilibrium is shifted by deprotonation of the thiol. The selenide analog of GSH, GSeH, has a low selenol pKa value and enhanced reactivity in both the diselenide and the selenide forms [18]. The diselenide is very reactive with thiolates and the selenide reacts rapidly with disulfides, even though a diselenide bond is intrinsically more stable than a disulfide bond. Furthermore, thiols have also been attached to microspheres and used successfully to enhance the folding of proteins, compared to nonthiol containing microspheres [19].
Protein folding The traditional model system for protein folding is ribonuclease A (RNase A). Experiments were initiated with RNase A in either the reduced denatured form (rRNase A), which contains eight cysteine thiols, or in the scrambled form (sRNase A), which has a random distribution of its four disulfide bonds (Table 1) [20,21,22]. Neither form possesses appreciable enzymatic activity. Upon the addition of either RNase A form to the folding mixture, the RNase A slowly regains enzymatic activity, usually over hours. A plot of enzymatic activity versus time is then fit to a single exponential, A(1 ekt), with A being the enzymatic activity observed at long times and k being the first-order folding rate constant (Figure 4). The value of A is also called the yield when expressed as a percent of the activity obtained if all the reduced or scrambled RNase A folded only to native RNase A. For comparing reagents, the standard conditions are 1 mM GSSG and 0.2 mM GSH. Most of the reagents have been compared to the standard conditions, though the pH and temperature varied somewhat [23]. Several of the folding catalysts have also been used to fold additional proteins in vitro and in vivo. BMC was shown to increase the in vivo folding yield of a secreted acid phosphatase in yeast threefold upon addition to the growth medium [10]. A more recent report using a fractional factorial screen showed that BMC performed better than both GSH/GSSG and dithiothreitol, DTT, at folding a range of disulfide-containing proteins in terms of
Table 1 Folding of reduced (r) or scrambled (s) RNase Aa Reagent Linear CXXC Cyclic CXXC photo CXXC cis/trans CXC or CGC BMC Aromatic 1 Aromatic 2 GSeH GSH a
Thiol pKa na 7.3, 9.6 na 8.7, 9.8 8.3, 9.9 6.6 5.7 5.3 (GSeH) 8.7
E80 (mV) (disulfide) 190 130 147/201 167 240 220 170 407 250
pH, T (8C), RNase A 7.4, 7.4, 7.4, 7.6, 7.6, 7.0, 7.0, 8.0,
20, 30, 30, 30, 30, 25, 25, 25,
r r r s s s s r
k (reagent/GSH) Twofold Twofold Approx. 1 Onefold to twofold Approx. 1 Sevenfold 12-fold Twofold
Yield (reagent/GSH) Approx. same 95%/83% 101%/77% 60%/20% 90%/40% Approx. same Approx. same Approx. same
Italics represent the best guess from the folding curves provided while na means that the data were not available.
www.sciencedirect.com
Current Opinion in Chemical Biology 2008, 12:740–745
744 Model Systems
Figure 4
Folding of scrambled RNase A (0.025 mM) at 25 8C and pH 7.7 in the presence of 1.0 mM GSH and 0.2 mM GSSG (standard conditions). The enzymatic activity was followed at discrete time points and fit to A(1 ekt), A and k values correspond to 91% yield and 0.008 min1, respectively.
yield [24]. Even at low concentrations, BMC increased the in vitro folding yield and rate. The in vitro folding rate of rRNase A by DTTox is enhanced twofold by the addition of 50 micromolar BMC [25]. At 10 micromolar BMC, the in vitro folding rate of proinsulin in the presence of GSH and GSSG increased only slightly but the yield of active protein went up from 20 to 40% [26]. Higher concentrations of BMC decreased the yield. It was concluded that BMC works at stages that affect aggregation but not folding rate. BMC could also be added to the growth medium of E. coli (2–50 micromolar) to increase the production of proinsulin by 60%. Again, higher concentrations were disadvantageous. Aromatic disulfides have been used previously for the rapid oxidation of thiols to disulfides, but until recently they have not been used for oxidative protein folding [27]. A series of aromatic thiols with a range of thiol pKa values were synthesized and tested for their ability to fold sRNase A. The effects of thiol pKa, thiol concentration, and solution pH on the folding rates were determined, and a general equation was obtained to predict folding rates [28]. Ultimately, a lower thiol pKa led to a faster folding rate under optimal condition, but higher thiol concentrations were required to reach optimal conditions. A thiol pKa approximately 1 unit less than the pH was viewed as optimal. Aromatic thiols were also demonstrated to enhance the folding rate of scrambled RNase A by threefold in the presence of PDI, though factors other than thiol pKa appeared to be important [29]. In addition, aromatic thiols were shown to improve the yield from 70 to 85% and the folding rate for lysozyme up to 11fold compared to glutathione [30]. Furthermore, an aromatic thiol, thiosalicylic acid, folded a single-chain antiCurrent Opinion in Chemical Biology 2008, 12:740–745
body fragment twice as fast as glutathione and increased the yield from 11 to 19% at pH 7 [31]. Immobilized Ellman’s reagent, an aromatic disulfide also known as CLEAR-OX, improved the folding yield of conotoxins containing two disulfide bonds by up to 10-fold or 20-fold, especially at high peptide concentrations, and improved the yield of conotoxins containing three disulfide bonds from 21 to 32% over those obtained in solution [32,33]. The effect of GSeSeG on the folding of bovine pancreatic trypsin inhibitor (BPTI) was more dramatic than that observed with RNase A [18]. At short times the amount of native protein produced was similar for both GSSG and GSeSeG, but at longer times GSeSeG provided significantly more native protein. Presumably, with GSeSeG the last native disulfide bond was formed via direct oxidation, as air was reported to enhance the reaction rate, whereas with GSSG it was formed by intramolecular rearrangement of disulfide bonds within the protein followed by oxidation [34]. Recently, substoichiometric amounts of GSeSeG have been shown to enhance RNase A folding relative to GSSG by catalyzing the oxidation folding of proteins with oxygen [35].
Conclusion In general, the small-molecule dithiols presented above will improve the yield of active protein. Aromatic thiols and selenols will enhance the protein folding rates. However, neither the aromatic thiols nor the selenols have been tested under conditions where the yield of active protein is low using standard conditions. The expectation is that small-molecule catalysts of oxidative protein folding will only continue to improve as more and more of the properties of PDI are incorporated into a single small molecule. These new catalysts will enhance the production of protein both in vivo and in vitro. As the link between oxidative protein folding and diseases is clarified, their in vivo applications will expand.
Acknowledgements The author would like to thank Drs David Chatfield, Fenfei Leng, and Alex Mebel for their helpful suggestions. Financial support from the National Science Foundation (CHE-0342167 to WJL) is gratefully acknowledged.
References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as: of special interest of outstanding interest 1.
Martelli PL, Fariselli P, Casadio R: Prediction of disulfide-bonded cysteines in proteomes with a hidden neural network. Proteomics 2004, 4:1665-1671.
2.
Nakamura T, Lipton SA: Emerging roles of S-nitrosylation in protein misfolding and neurodegenerative diseases. Antioxid Redox Signal 2008, 10:87-101.
3.
Sahdev S, Khattar SK, Saini KS: Production of active eukaryotic proteins through bacterial expression systems: a review of the existing biotechnology strategies. Mol Cell Biochem 2008, 307:249-264. www.sciencedirect.com
Small-molecule catalysts of oxidative protein folding Lees 745
4.
Kersteen EA, Raines RT: Catalysis of protein folding by protein disulfide isomerase and small-molecule mimics. Antioxid Redox Signal 2003, 5:413-424. A nice review of the properties and function of PDI.
19. Woycechowsky KJ, Hook BA, Raines RT: Catalysis of protein folding by an immobilized small-molecule dithiol. Biotechnol Progr 2003, 19:1307-1314.
5.
20. Crook EM, Mathias AP, Rabin BR: Spectrophotometric assay of bovine pancreatic ribonuclease by the use of cytidine 20 :30 phosphate. Biochem J 1960, 74:234-238.
6.
21. Konishi Y, Scheraga HA: Regeneration of ribonuclease A from the reduced protein. 1. Conformational analysis of the intermediates by measurements of enzymic activity, optical density, and optical rotation. Biochemistry 1980, 19:1308-1316.
Tian G, Xiang S, Noiva R, Lennarz WJ, Schindelin H: The crystal structure of yeast protein disulfide isomerase suggests cooperativity between its active sites. Cell (Cambridge, MA, USA) 2006, 124:61-73. The crystal structure of PDI.
7.
8.
9.
Lundstroem J, Holmgren A: Determination of the reduction– oxidation potential of the thioredoxin-like domains of protein disulfide-isomerase from the equilibrium with glutathione and thioredoxin. Biochemistry 1993, 32:6649-6655. Hawkins HC, Freedman RB: The reactivities and ionization properties of the active-site dithiol groups of mammalian protein disulfide-isomerase. Biochem J 1991, 275:335-339. Darby N, Creighton TE: Characterization of the active site cysteine residues of the thioredoxin-like domains of protein disulfide isomerase. Biochemistry 1995, 34:16770-16780. Walker KW, Gilbert HF: Scanning and escape during proteindisulfide isomerase-assisted protein folding. J Biol Chem 1997, 272:8845-8848.
10. Woycechowsky KJ, Wittrup KD, Raines RT: A small-molecule catalyst of protein folding in vitro and in vivo. Chem Biol 1999, 6:871-879. This paper describes the first use of a nonpeptide small-molecule dithiol for the folding of RNase A and demonstrates the successful in vivo application of a small-molecule catalyst of oxidative protein folding. 11. Appenzeller-Herzog C, Ellgaard L: In vivo reduction–oxidation state of protein disulfide isomerase: the two active sites independently occur in the reduced and oxidized forms. Antioxid Redox Signal 2008, 10:55-64. 12. Moroder L, Besse D, Musiol H-J, Rudolph-Boehner S, Sideler F: Oxidative folding of cystine-rich peptides versus regioselective cysteine pairing strategies. Biopolymers 1996, 40:207-234. 13. Cabrele C, Fiori S, Pegoraro S, Moroder L: Redox-active cyclic bis(cysteinyl)peptides as catalysts for in vitro oxidative protein folding. Chem Biol 2002, 9:731-740. This paper depicts the use of cyclic CXXC peptides that better mimic the properties of PDI than the original linear CXXC peptides. 14. Cabrele C, Cattani-Scholz A, Renner C, Behrendt R, Oesterhelt D, Moroder L: Photomodulation of the redox and folding adjuvant properties of bis(cysteinyl) peptides. Eur J Org Chem 2002, 2002:2144-2150. 15. Ookura T, Kainuma K, Kim H-j, Otaka A, Fujii N, Kawamura Y: Active site peptides with CXXC motif on MAP-resin can mimic protein disulfide isomerase activity. Biochem Biophys Res Commun 1995, 213:746-751. 16. Woycechowsky KJ, Raines RT: The CXC motif: a functional mimic of protein disulfide isomerase. Biochemistry 2003, 42:5387-5394. 17. DeCollo TV, Lees WJ: Effects of aromatic thiols on thiol– disulfide interchange reactions that occur during protein folding. J Org Chem 2001, 66:4244-4249. 18. Beld J, Woycechowsky KJ, Hilvert D: Selenoglutathione: efficient oxidative protein folding by a diselenide. Biochemistry 2007, 46:5382-5390. This account portrays the first use of the more reactive diselenides to fold proteins.
www.sciencedirect.com
22. Gough JD, Williams RH Jr, Donofrio AE, Lees WJ: Folding disulfide-containing proteins faster with an aromatic thiol. J Am Chem Soc 2002, 124:3885-3892. This paper describes the first use of an aromatic thiol for the folding of disulfide-containing proteins. 23. Lyles MM, Gilbert HF: Catalysis of the oxidative folding of ribonuclease A by protein disulfide isomerase: dependence of the rate on the composition of the redox buffer. Biochemistry 1991, 30:613-619. 24. Willis MS, Hogan JK, Prabhakar P, Liu X, Tsai K, Wei Y, Fox T: Investigation of protein refolding using a fractional factorial screen: a study of reagent effects and interactions. Protein Sci 2005, 14:1818-1826. 25. Fink M, Nieves P, Chang S, Narayan M: Non-redox-active smallmolecules can accelerate oxidative protein folding by novel mechanisms. Biophys Chem 2008, 132:104-109. 26. Winter J, Lilie H, Rudolph R: Recombinant expression and in vitro folding of proinsulin are stimulated by the synthetic dithiol Vectrase-P. FEMS Microbiol Lett 2002, 213:225-230. 27. Annis I, Chen L, Barany G: Novel solid-phase reagents for facile formation of intramolecular disulfide bridges in peptides under mild conditions. J Am Chem Soc 1998, 120:7226-7238. 28. Gough JD, Lees WJ: Effects of redox buffer properties on the folding of a disulfide-containing protein: dependence upon pH, thiol pKa, and thiol concentration. J Biotechnol 2005, 115:279-290. This is a more detailed account of the use of aromatic thiols to fold proteins and provides some understanding of the important factors. 29. Gough JD, Lees WJ: Increased catalytic activity of protein disulfide isomerase using aromatic thiol based redox buffers. Bioorg Med Chem Lett 2005, 15:777-781. 30. Gurbhele-Tupkar MC, Perez LR, Silva Y, Lees WJ: Rate enhancement of the oxidative folding of lysozyme by the use of aromatic thiol containing redox buffers. Bioorg Med Chem 2008, 16:2579-2590. 31. Patil G, Rudolph R, Lange C: In vitro-refolding of a single-chain Fv fragment in the presence of heteroaromatic thiols. J Biotechnol 2008, 134:218-221. 32. Darlak K, Long DW, Czerwinski A, Darlak M, Valenzuela F, Spatola AF, Barany G: Facile preparation of disulfide-bridged peptides using the polymer-supported oxidant CLEAR-OX. J Pept Res 2004, 63:303-312. 33. Green BR, Bulaj G: Oxidative folding of conotoxins in immobilized systems. Protein Pept Lett 2006, 13:67-70. 34. Kibria FM, Lees WJ: Balancing conformational and oxidative kinetic traps during the folding of bovine pancreatic trypsin inhibitor (BPTI) with glutathione and glutathione disulfide. J Am Chem Soc 2008, 130:796-797. 35. Beld J, Woycechowsky KJ, Hilvert D: Catalysis of oxidative protein folding by small-molecule diselenides. Biochemistry 2008, 47:6985-6987.
Current Opinion in Chemical Biology 2008, 12:740–745