Norman Metanis

Traceless Ligation of Cysteine Peptides Using Selective Deselenization

The synthesis of proteins with a fully native sequence is an ongoing challenge in protein chemistry. Native chemical ligation (NCL) approaches have proven to be generally applicable where cysteine (Cys) residues are appropriately positioned,[1,2] however, the synthesis of many proteins often require ligation at non-Cys sites in the polypeptide sequence.[3-7] Previously, we introduced a reductive strategy for ligation at Ala sites[7] based on global desulfurization of Cys[8] that has found widespread utility for the synthesis of complex proteins by NCL.[9,10] Selective desulfurization can be affected by both Rainey Ni and Pd/C/H2[7] and, more recently, by the radical initiator VA-044 in combination with the water soluble phosphine TCEP.[11] However, since these conditions result in global desulfurization of all thiols in the protein, the method requires protection and deprotection of all other Cys residues in the native sequence.[6a,12] These additional steps complicate the synthesis of larger polypeptides and limit the use of natural Cys residues for ligation.[13]

Natural and synthetic selenoproteins

Once considered highly toxic, the element selenium is now recognized as a micronutrient essential for human health. It is inserted co-translationally into many proteins as the non-canonical amino acid selenocysteine, providing the resulting selenoprotein molecules with a range of valuable redox properties; selenocysteine is also increasingly exploited as a structural and mechanistic probe in synthetic peptides and proteins. Here we review topical investigations into the preparation and characterization of natural and artificial selenoproteins. Such molecules are uniquely suited as tools for protein chemistry and as a test bed for studying novel catalytic activities.

Strategic Use of Non‐Native Diselenide Bridges to Steer Oxidative Protein Folding

Oxidation of cysteine thiols to disulfides is a widespread posttranslational modification of exported proteins, including many of pharmaceutical interest. For disulfide-rich proteins, many different disulfide-bonded species are possible, and the pathway from a reduced unfolded polypeptide chain to the fully oxidized, native conformation may entail multiple oxidation, reduction, and rearrangement steps. Not surprisingly, formation of scrambled disulfide isomers or accumulation of kinetically and conformationally trapped intermediates often limits the rates and yields of oxidative protein folding. Biotechnological production of such proteins would consequently benefit from strategies that improve folding efficiency. We recently showed that redox buffers containing selenoglutathione (GSeSeG), which is a diselenide analogue of glutathione (GSSG), promote the oxidative folding of a wide range of proteins. Diselenides are thermodynamically more stable than disulfides but react more rapidly with thiols owing to the higher polarizability of selenium compared to sulfur. The greater acidity of selenols versus thiols (DpKa 3) further enhances their reactivity. For example, deprotonated selenolate anions are excellent nucleophiles that are capable of catalyzing disulfide bond shuffling. As a consequence, diselenide-containing redox buffers typically afford higher rates and yields of native protein than conventional glutathione redox buffers over a broader pH range and at substantially lower concentrations. Because selenols are rapidly oxidized by molecular oxygen, GSeSeG can even function as a true catalyst for aerobic folding reactions. The utility of GSeSeG is particularly striking for proteins, such as bovine pancreatic trypsin inhibitor (BPTI), whose folding mechanisms are dominated by kinetically trapped species. Diselenide-containing redox buffers rescue such intermediates, much like the natural enzyme protein disulfide isomerase (PDI), by accelerating rate-limiting disulfide bond rearrangements. While intermolecular reactions of diselenides and reduced protein can be quite efficient, intramolecular reactions would be expected to provide even greater control over oxidative folding pathways. In fact, diselenides have been widely used as isosteric replacements of native disulfides in cysteine-rich proteins, such as endothelins and conotoxins, to increase folding rates and yields while maintaining biological activity. Because diselenides have significantly lower reduction potentials than disulfides, their formation is favored thermodynamically, providing a powerful intramolecular constraint that ensures regioselective formation of correct cross-links. Moreover, the diselenides increase the stability of these molecules to reduction and disulfide scrambling. In contrast, non-native diselenide cross-links are useful for generating thermodynamically stable models of kinetically unstable folding intermediates. For example, non-productive conformations of the 18-amino acid peptide apamin have been successfully trapped by this approach. For proteins more stable than apamin, however, incorrectly cross-linked intermediates need not necessarily be kinetic dead ends, provided sufficient conformational folding energy is available to override the thermodynamic benefit associated with diselenide formation. We hypothesized that it might even be possible to exploit non-native diselenides in such cases to bias early folding events and strategically direct folding toward a specific productive route. BPTI, one of the first proteins for which a detailed folding pathway was deduced, 14] is 58 amino acids long and stabilized by three disulfide bonds, which are designated by the positions of the paired cysteines, i.e. [5–55; 14–38; 30–51] (Figure 1a). The reduced protein folds in the presence of glutathione via a bifurcated pathway involving a small number of discrete intermediates that contain one and two native disulfide bonds (Figure 1b). Roughly half of the reduced BPTI molecules reach the native state via an intermediate (N’) that lacks the 5–55 disulfide; the other half are trapped as a stable, native-like species (N*) that lacks the 30–51 disulfide. Formation of the fully oxidized native state requires partial unfolding and subsequent rate-limiting disulfide bond rearrangements of these intermediates. At neutral pH, N* is stable for weeks, greatly limiting folding efficiency. 14] In the early stages of BPTI folding, a broad spectrum of one-disulfide intermediates is initially formed, but the population quickly becomes dominated by species containing native 5–55 and 30–51 disulfides. We speculated that replacing both Cys5 and Cys14 with selenocysteine (usually abbreviated as Sec or U) would perturb this normal steadystate distribution owing to preferential formation of the nonnative 5–14 cross-link. Biasing the system in this way might further influence the set of two-disulfide intermediates that arise and thus open alternative folding channels to the fully oxidized protein (Figure 1 b). For example, the 5–14 diselenide would be expected to stabilize the non-native [5–14; 30– 51] intermediate, which has been detected at low levels in [*] Dr. N. Metanis, Prof. Dr. D. Hilvert Laboratory of Organic Chemistry, ETH Z rich 8093 Z rich (Switzerland) E-mail: hilvert@org.chem.ethz.ch

Chemical Synthesis of Proteins with Non‐Strategically Placed Cysteines Using Selenazolidine and Selective Deselenization

Although native chemical ligation has enabled the synthesis of hundreds of proteins, not all proteins are accessible through typical ligation conditions. The challenging protein, 125-residue human phosphohistidine phosphatase 1 (PHPT1), has three cysteines near the C-terminus, which are not strategically placed for ligation. Herein, we report the first sequential native chemical ligation/deselenization reaction. PHPT1 was prepared from three unprotected peptide segments using two ligation reactions at cysteine and alanine junctions. Selenazolidine was utilized as a masked precursor for N-terminal selenocysteine in the middle segment, and, following ligation, deselenization provided the native alanine residue. This approach was used to synthesize both the wild-type PHPT1 and an analogue in which the active-site histidine was substituted with the unnatural and isosteric amino acid β-thienyl-l-alanine. The activity of both proteins was studied and compared, providing insights into the enzyme active site.

Selenium and Selenocysteine in Protein Chemistry

Selenocysteine, the selenium-containing analogue of cysteine, is the twenty-first proteinogenic amino acid. Since its discovery almost fifty years ago, it has been exploited in unnatural systems even more often than in natural systems. Selenocysteine chemistry has attracted the attention of many chemists in the field of chemical biology owing to its high reactivity and resulting potential for various applications such as chemical modification, chemical protein (semi)synthesis, and protein folding, to name a few. In this Minireview, we will focus on the chemistry of selenium and selenocysteine and their utility in protein chemistry.

Revisiting ligation at selenomethionine: Insights into native chemical ligation at selenocysteine and homoselenocysteine

Selenomethionine (Sem) has been incorporated recombinantly into proteins many times to elucidate their structure and function. In this paper, we revisit incorporation via chemical protein synthesis to shed light on the mechanism of native chemical ligation. The effect of chalcogen position on ligation is investigated, and selenium-containing peptide ligation is optimized. Additionally, selective methylation is performed on selenolates in a peptide in the presence of unprotected thiols.

The Chemistry of Selenocysteine in Proteins

While the existence of the rare 21st proteinogenic amino acid, selenocysteine (Sec, U) in cellular proteins has been known for over 40 years, recent advances in peptide chemistry have supported its importance not only in the biological function of natural selenoproteins, but also in synthetic systems. Besides its obvious applications in the synthesis of selenoproteins, fundamental differences between the chemistry of cysteine’s thiol and Sec’s selenol moieties have added a surprising number of technologies to the protein chemist’s toolbox. In this chapter, we discuss Sec’s impact on chemical protein synthesis, folding of challenging disulfide-rich proteins, and the chemistry of the little-understood selenoproteins, SEP15 and SELM. Additional important aspects on Sec will be the subjects of other chapters in this book.

A combinatorial approach for the synthesis of multi-phosphorylated peptides: new tool for studying phosphorylation patterns in proteins

Phosphorylation of proteins at multiple sites creates different phosphorylation patterns that are essential for their biological activity. For example, such patterns contribute to the redirection of signalling to alternative pathways. Multi-phosphorylated peptides are excellent tools to systematically study the impact of unique phosphorylation patterns on signalling, but their synthesis is extremely difficult. Here we present an efficient and general method for the synthesis of multi-phosphorylated peptides, using a combination of different tailor-made coupling conditions. The method was demonstrated for the synthesis of a library of Rhodopsin C terminal peptides with distinct phosphorylation patterns containing up to seven phosphorylated Ser (pSer) and Thr (pThr) residues in close proximity to one another. Our method can be used to synthesize peptides incorporating multiple phosphorylated amino acids at high efficiency. It does not require any special expertise and can be performed in any standard peptide laboratory. This approach opens the way for quantitative mechanistic studies of phosphorylation patterns and their biological roles.

Peptide fibrils as monomer storage of the covalent HIV‐1 integrase inhibitor

We have recently reported the covalent inhibition of HIV‐1 integrase by an N‐terminal succinimide‐modified lens epithelium‐derived growth factor (361–370) peptide. We also showed that this peptide is proteolytically stable. Here, we show that this inhibitor is stored as fibrils that serve as a stock for the inhibitory monomers. The fibrils increase the local concentration of the peptide at the target protein. When the monomers bind integrase, the equilibrium between the fibrils and their monomers shifts towards the formation of peptide monomers. The combination of fibril formation and subsequent proteolytic stability of the peptide may bring to new strategy for developing therapeutic agents. Copyright

Covalent Inhibition of HIV‐1 Integrase by N‐Succinimidyl Peptides

We present a new approach for the covalent inhibition of HIV‐1 integrase (IN) by an LEDGF/p75‐derived peptide modified with an N‐terminal succinimide group. The covalent inhibition is mediated by direct binding of the succinimide to the amine group of a lysine residue in IN. The peptide serves as a specific recognition sequence for the target protein, while the succinimide serves as the binding moiety. The combination of a readily synthesizable peptide precursor with easy and efficient binding to the target protein makes this approach a promising new strategy for designing lead compounds.