Novel Methods for the Chemical Synthesis of Insulin Superfamily

Aug 22, 2017 - In 2005, he joined the Florey Institute of Neuroscience and Mental Health in Melbourne, Australia, where his research interests involve...
0 downloads 10 Views 3MB Size
Article pubs.acs.org/accounts

Novel Methods for the Chemical Synthesis of Insulin Superfamily Peptides and of Analogues Containing Disulfide Isosteres Published as part of the Accounts of Chemical Research special issue “Chemical Biology of Peptides”. Mohammed Akhter Hossain* and John D. Wade* The Florey Institute of Neuroscience and Mental Health and School of Chemistry, University of Melbourne, Melbourne, Victoria 3010, Australia CONSPECTUS: The insulin superfamily of peptides is ubiquitous within vertebrates and invertebrates and is characterized by the presence of a set of three disulfide bonds in a unique disposition. With the exception of insulin-like growth factors I and II, which are single chain peptides, the remaining 8 members of the human insulin superfamily are two-chain peptides containing one intramolecular and two intermolecular disulfide bridges. These structural features have long made the chemical synthesis of the peptides a considerable challenge, in particular, including their correct disulfide bond pairing and formation. However, they have also afforded the opportunity to develop modern solid phase synthesis methods for the preparation of such peptides that incorporate novel or improved chemical methods for the controlled introduction of both disulfide bonds and their surrogates, both during and after peptide chain assembly. In turn, this has enabled a detailed probing of the structure and function relationship of this small but complex superfamily of peptides. After initially using and subsequently identifying significant limitations of the approach of simultaneous random chain combination and oxidative folding, our laboratory undertook to develop robust chemical synthesis strategies in concert with orthogonal cysteine S-protecting groups and corresponding regioselective disulfide bond formation. These have included the separate synthesis of each of the two chains or of the two chains linked by an artificial C-peptide that is removed following postoxidative folding. These, in turn, have enabled an increased ease of acquisition in a good yield of not only members of human insulin superfamily but other insulin-like peptides. Importantly, these successful methods have enabled, for the first time, a detailed analysis of the role that the disulfide bonds play in the structure and function of such peptides. This was achieved by selective removal of the disulfide bonds or by the judicious insertion of disulfide isosteres that possess structurally subtle variations in bond length, hydrophobicity, and angle. These include lactam, dicarba, and cystathionine, each of which has required modifications to the peptide synthesis protocols for their successful placement within the peptides. Together, these synthesis improvements and the novel chemical developments of cysteine/cystine analogues have greatly aided in the development of novel insulin-like peptide (INSL) analogues, principally with intra-A-chain disulfide isosteres, possessing not only improved functional properties such as increased receptor selectivity but also, with one important and unexpected exception, greater in vivo half-lives due to stability against disulfide reductases. Such analogues greatly will aid further biochemical and pharmacological analyses to delineate the structure−function relationships of INSLs and also future potential drug development.



INTRODUCTION The then unexpected finding that the glycemic hormone, insulin, consists of not one but two peptide chains, A and B, linked by three disulfide bonds in a unique disposition was a milestone in protein chemistry and biochemistry.1 The disulfide crossbridging comprising an intra-A-chain and two interchain disulfide bonds was soon recognized to represent a signature structural feature (Figure 1A).2 Early protein chemistry studies followed by genomic or expressed sequence tag database sequencing has led to the continuing identification of numerous insulin-like peptides in both vertebrates and invertebrates.3−5 In Homo sapiens, the insulin superfamily consists of ten members: insulin, insulin-like growth factors (IGF) I and II, and a relaxin © 2017 American Chemical Society

subfamily that comprises relaxin-1, -2, and -3 and four other insulin-like peptides (INSL 3, 4, 5, and 6).3,5,6 IGFs I and II are unique in that they both possess a single peptide chain structure that is cross-linked by three disulfide bonds in the same disposition as for insulin. Whereas insulin and IGFs I and II interact with tyrosine kinase receptors, the members of the relaxin subfamily interact with G-protein-coupled receptors known as relaxin family peptide receptors.7 The diversity of biological activity within the broad superfamily is striking, ranging from metabolic control through to gubernacular Received: June 9, 2017 Published: August 22, 2017 2116

DOI: 10.1021/acs.accounts.7b00288 Acc. Chem. Res. 2017, 50, 2116−2127

Article

Accounts of Chemical Research

Figure 1. (A) Primary structure of human insulin showing the disposition of the signature disulfide bond pairings. (B) The tertiary structure of human insulin.

Account highlights our development of solid phase peptide synthesis (SPPS) protocols that have allowed us to achieve these goals together with our observations on the effect of such bonds on the chemical biology of these peptides.

development, cardiovascular biology, and collagen homeostasis.3,7 Much of our knowledge of the biology of these peptides has stemmed from their availability via their chemical synthesis, which has progressively developed from the original preparation of bovine insulin nearly six decades ago.8,9 A comparison of the known tertiary structures of several human INSLs together with insulin highlights a remarkable degree of structural similarity, which indicates that the biological diversity is principally conferred by the primary structures themselves.10 An INSL A-chain characteristically consists of terminal α-helices linked by a noncanonical turn via the intrachain disulfide bond into a nearantiparallel alignment. The B-chain contains a central α-helix with interchain disulfide bridges at each end (Figure 1B).11,12 To date, there is no evidence of a functional role for any of the disulfide bonds in an INSL indicating that these act to mediate precursor folding and maintain the mature INSL in a biologically active conformation. Removal of any one of the three disulfide bonds from insulin causes drastic loss of both biological activity and secondary structure.13 Clearly, each of the three disulfide bonds is irreplaceable other than with an isosteric bond. Such replacement is attractive given that minor changes in length, bond angle, or hydrophilicity of the cross-link can introduce subtle variations in function through modified receptor interactions. Importantly, these will also increase the metabolic half-life of the peptide as cellular disulfide reduction is a significant component of the in vivo protein degradation pathway.14 However, an examination of the effect of replacement of disulfides within an INSL has been limited until recent years by the inability to both efficiently chemically assemble these complex peptides and adapt or develop novel chemistry to successfully introduce cystine isosteres. Consequently, this



CHEMICAL SYNTHESIS OF INSULIN-LIKE PEPTIDES

Random Chain Combination

Bovine insulin was first assembled via separate solution phase synthesis of the S-reduced A- and B-chains and their combination in solution at high pH to spontaneously oxidatively fold and generate the native three cystine bonds.8 Improved yields of folding were obtained when the more readily handled and purified S-sulfonated forms of the A- and B-chains were used in the presence of reducing agent (Figure 2A).9 These early results showed that the primary structure of the peptide chains was the key determinant for the correct alignment and subsequent cystine formation, itself remarkable given that the number of statistically possible disulfide heterodimers of the two insulin chains is 12. This approach enabled combination of Boc-solid phase-synthesized chains of bovine insulin15 and, in our laboratory, porcine relaxin and a human relaxin-2 analogue.16,17 We later observed that truncated or certain point-mutated chains were generally resistant to combination thus necessitating the development of alternative synthesis methods. Regioselective Disulfide Bond Formation

The advent of orthogonal cysteine S-thiol protecting groups led to the development of effective approaches to regioselectively form each of the disulfide bonds of insulin following a solution phase fragment synthesis approach.18 These were soon adapted for the SPPS of bombyxin-IV, insulin, and human relaxin-2.19−21 The bombyxin-IV synthesis strategy was notably effective and 2117

DOI: 10.1021/acs.accounts.7b00288 Acc. Chem. Res. 2017, 50, 2116−2127

Article

Accounts of Chemical Research

Figure 2. (A) Preparation of insulin via random oxidative folding of the individual S-sulfonated A- and B-chains. (B) INSL preparation via regioselective disulfide bond formation within the S-protected A- and B-chains.

after which the solubilizing “tail” was simply removed from the resin by brief aqueous sodium hydroxide treatment to provide insulin glargine in greater overall yield.37 Despite the effectiveness of this strategy, the multiple steps required for both the separate chain syntheses and the subsequent individual disulfide bond formation invariably lead to overall modest yields of purified target peptide (usually 10% relative to the limiting B-chain). While not critical for the preparation of INSL analogues for structural and functional study, further improvements are desirable for the chemical synthesis of such peptides, in particular, still higher yields of SPPS of the chains, novel S-protection that affords additional levels of orthogonality, and improved or alternative chemically directed disulfide formation that avoids damage to other residues such as Trp and Tyr. The development of photolabile S-protecting groups provides opportunities for both “one pot” and regioselective disulfide bond formation. It also enables the replacement of an often lowyielding oxidative cleavage of two S-Acm-protected cysteine residues by iodine that is often associated with the formation of side products. While several photolabile S-protecting groups including the S-nitrobenzyl38 are available, these often require protracted conditions for their removal. The S-2-nitroveratryl (S-oNv) group39 possesses properties that augur well for its use in peptide chemistry, in particular, its ambient removal at long wavelength (350 nm), which is less destructive to amino acids such as Trp. Its use with a cysteine “partner” such as S-Pyr, allows efficient thiolysis to occur after photolytic cleavage and also reduces unwanted disulfide shuffling (Figure 4A). Fmoc-Cys(oNv)-OH was synthesized in good yield and used for the assembly of human insulin (Figure 4B).40 Use of an iso-acyl dipeptide within the A-chain enhanced its solubility in

employed only Fmoc-based SPPS for the two chains that were selectively S-protected with a combination of S-trityl (Trt), S-acetamidomethyl (Acm), and tert-butyl (tBu) groups. After A-chain intramolecular disulfide bond formation by oxidation of the two free thiols with 2,2′-dipyridyldisulfide (DPDS), treatment of the resulting peptide intermediate with DPDS in trifluoromethanesulfonic acid (TFMSA) led to the displacement of the C-terminal Cys(tBu) group and corresponding formation of Cys(pyridinylsulfenyl; Pyr) derivative. This then reacted with the free thiol of the C-terminal Cys of the B-chain to form the first intermolecular disulfide. The second and final intermolecular disulfide was generated by iodolysis of the pair of Cys(Acm) residues (Figure 2B). We subsequently found that this strategy worked well for the first acquisition of human relaxin-3 (H3 relaxin), which was previously shown to be refractive to the random chain refolding approach,22 and enabled its tertiary structure determination by NMR spectroscopy.23 The synthetic strategy was used to prepare other members of the relaxin subfamily of peptides including INSL3, INSL5, and R3/I5 and to determine their tertiary structures,10,24−26 as well as relaxin-2 from other species including horse and mouse.27−29 Numerous analogues of relaxin subfamily peptides, including versions that are size-reduced and site-specifically labeled with europium or fluorescent tags for receptor interaction studies have also been obtained.30−35 A successful assembly of Drosophila insulin-like peptide 2 was also achieved.36 A modified strategy was also developed in which the C-terminus of the A-chain of insulin glargine was assembled on the solid support containing penta-Lys attached by a baselabile 4-hydroxymethylbenzoic acid (HMBA) linker.37 This greatly improved the solubility of A-chain and its subsequent purification and disulfide pairing with the B-chain (Figure 3) 2118

DOI: 10.1021/acs.accounts.7b00288 Acc. Chem. Res. 2017, 50, 2116−2127

Article

Accounts of Chemical Research

Figure 3. RP-HPLC of synthetic insulin glargine A-chain without (A) and with (B) the C-terminal penta-Lys “solubilizing tail”. (C) Comparison of solubility of the two peptides in H2O (1 mg/mL). Adapted from ref 37. Copyright 2009 American Chemical Society.

aqueous media.41 After cleavage of the A-chain depsipeptide, formation of the A6−A11 intrachain disulfide bond and combination of the A- and B-chains by our established methods, S-Pyr functionalization at CysB7 was followed by photolysis that was complete after 45 min. Conversion of the depsipeptide to the native insulin was accomplished by a pH adjustment from 5.5 to 7.5. The final characterized product (7% yield relative to B-chain) had native affinity for the insulin receptor (Figure 4C−E). This approach provides scope for optimization particularly in combination with wavelength-orthogonal photolabile S-protecting groups.42

be successful for efficient folding of highly modified insulins46 such as those containing pendant molecules or cystine isosteres. We developed a related directed biomimetic folding strategy47 that can also be combined with modern native chemical ligation synthesis of precursor peptide intermediates and which provides a basis for the scalable synthesis of insulin and of analogues that may be unobtainable by our previously established methods. This employs a novel dimedone-based linker together with an HMBA-based linker to synthesize a heterodimeric peptide (thionin analogue) as well as peptide−peptide conjugates (Figure 5).47 This bis-linker tether method is compatible with both Fmoc- and Boc-SPPS. Since this strategy reduced the number of intermediate and purification steps, the peptides were obtained in good overall yields (>24%).47 This method will be applicable to the preparation of two-chain cystine-rich peptides and proteins including INSLs.

Biomimetic Folding and Disulfide Bond Formation

INSLs are produced in vivo on the ribosome as pre-prohormones consisting of a leader presequence, the B-chain, a connecting C-peptide and the A-chain. After folding and disulfide bond formation, the pre- and C-peptides are enzymatically excised to yield the native INSL.14 Chemical methods have been developed that mimic this folding pathway. Synthetic insulin Lispro ([LysB28,ProB29]-human insulin; “KP-insulin”) was prepared using a nondirected disulfide bond formation strategy in which the A- and B-chains of insulin are directly connected via an ester bond between the β-hydroxyl group of ThrB30 and the γ-carboxyl group of GluA4.43 This surrogate proinsulin undergoes efficient folding and disulfide bond formation. The ThrB30− GluA4 ester linkage is then simply cleaved by saponification. A drawback of this method is the necessity for the presence of specifically colocated GluA4 and ThrB30 for tethering the two chains to enable pro-hormone-like folding. A DesDi insulin precursor44 has also been reported, which possesses an amide bond between the N-terminus of the A-chain and the C-terminus of a B-chain truncated to position 28 that is Pro substituted by Lys. After folding, the two-chain insulin analogue is obtained following Lys-C endoproteinase treatment. This method also has a limitation in that the template precursor is a B-chain-truncated insulin containing a non-native Lys at position 28. A related chain oxime-tethered biomimetic method for human insulin led to an overall yield of 20% based on purified A-chain.45 While these elegant nondirected biomimetic folding methods afford high overall yields (3−20%) of certain insulin analogues, they may not



DISULFIDE BOND ISOSTERES Although an excellent structural element, the disulfide bond is both chemically and metabolically labile and can undergo conversion to Cys residues by reduction or scrambling by thiolcontaining molecules such as glutathione or by nucleophilic or basic conditions.14 Disulfide bond substitution can prevent such modifications. Various isosteres have been investigated including diselenide, dicarba, lactam, thioether, and triazole bridges. Each generally confers serum stability although not always with retention of full activity, which likely reflects the differences in bond lengths and geometry. The cystine bond is 2.02 Å in length whereas the diselenium bond is 2.33 Å and the dicarba bond is considerably shorter at 1.54 Å (unsaturated) or 1.34 Å (saturated). The lactam (amide) bond is also short at approximately 1.47 Å depending upon the planarity of the bridge, while the cystathionine bond length, at 1.82 Å, is closest to that of the disulfide bond. The latter bond also closely approximates the geometry of the disulfide bond and thus is expected to cause minimal structural perturbations.48 Although diselenides and selenoethers are excellent disulfide mimics, there are safety concerns over the long-term use of selenocysteine-containing peptides. Despite having a slightly 2119

DOI: 10.1021/acs.accounts.7b00288 Acc. Chem. Res. 2017, 50, 2116−2127

Article

Accounts of Chemical Research

Figure 4. (A) Disulfide bond formation by photocleavage of the oNv protecting group and concomitant thiolysis by S-Pyr activation. (B) Insulin synthesis strategy. RP-HPLC (C) and MALDI-TOF MS (D) of synthetic human insulin. (E) Competition binding assay of synthetic human insulin with europium-labeled insulin to immunocaptured insulin receptor-A. Adapted from ref 40. Copyright 2014 Wiley-VCH Verlag GmbH & Co. KGaA.



different geometry and restricted rotation, triazoles have been successfully used as isosteric replacements of disulfides without loss of biological activity.49 However, there is also potential for adversely altering the three-dimensional structure with loss of peptide activity. Insulin analogues in which the A7−B7 interchain disulfide bond in human insulin was replaced with one of two triazole linkages (1,4 or 1,5)50 were found to have significantly unfolded structure and were completely inactive. Thus, maintaining a proper interchain orientation is critical for the structure and function of INSLs. Therefore, we focused upon methods for inserting other disulfide isosteres including the lactam, dicarba, and cystathionine bonds in order to prepare active INSL analogues.

CHEMICAL SYNTHESIS OF DISULFIDE ISOSTERES OF INSLs

Lactam-Containing INSLs

The incorporation of lactam bridges in peptides has been widely reported.51,52 We used a two-chain assembly strategy on a solid support for synthesizing a bis-lactam interchain-linked heterodimeric (ΔA25/26) INSL3 analogue (Figure 6).53 Briefly, the B-chain was assembled in which CysB10 and CysB22 were substituted with Lys(Mmt) and Lys(ivDde), respectively. After synthesis, on-resin removal of the ivDde group with 3% hydrazine in DMF was followed by lactam bond formation with Fmoc-Asp-OBut. After Nα-Fmoc-group removal, synthesis 2120

DOI: 10.1021/acs.accounts.7b00288 Acc. Chem. Res. 2017, 50, 2116−2127

Article

Accounts of Chemical Research

Figure 5. (A) Regioselective synthesis of thionin via bis-linker strategy. Conditions: (a) 1.5 equiv of DPDS, (b) 60 mM I2/HOAc, (c) 5% hydrazine. (B) RP-HPLC of peptide folding: crude reduced thionin analogue (A-bis-linker tether-B), 3A; purified oxidized thionin analogue (A-bis-linker tether-B) with one disulfide bond, 3B; purified oxidized thionin analogue (A-bis-linker tether-B) with two disulfide bonds, 3C; crude thionin analogue (A−B), 3D; cleaved bis-linker tether, 3E. Adapted from ref 47. Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA.

of the A-chain continued under forcing, microwave-enhanced conditions. The CysA11 residue was replaced with Asp(O-2-PhiPr). Both the O-2-PhiPr and Mmt protecting groups were removed

with 1% TFA in DCM after which the second interchain lactam bond (A11/B10) was formed by HTCU activation. Fmoc-SPPS continued through to residue 1 of the A-chain. 2121

DOI: 10.1021/acs.accounts.7b00288 Acc. Chem. Res. 2017, 50, 2116−2127

Article

Accounts of Chemical Research

in the production of the saturated dicarba bond. An alternative approach involves preprepared orthogonally protected diamino diacids as SPPS building blocks that enable on-resin lactamization to generate dicarba bonds.62 Microwave irradiation-mediated on-resin RCM was applied to the preparation of an intra-A-chain dicarba analogue of human relaxin-3 (Figure 7A).59 The A-chain was assembled by conventional SPPS in which the CysA10,15 residues were replaced with Hag and concluded with the retention of the Nα-Fmoc protecting group of the N-terminal residue. After dicarba bond formation (Figure 7B) and removal of the N-terminal Fmoc group, the resulting peptide was cleaved from the resin. Two peaks were identified by RP-HPLC each with the expected calculated molecular mass and corresponding to two isomers (cis/trans), which were separately combined with the B-chain (Figure 7C,D). CD spectroscopic analyses of the two isomers showed similar spectra to native human relaxin-3. Solution NMR spectroscopy and molecular modeling suggested that the major isomeric product had a global conformation that was close to the native relaxin-3 hormone and was the cis form, which more closely resembles the native disulfide bond. Both isomers were equipotent in binding to and activating the RXFP3 receptor. An intra-A-chain dicarba analogue of human INSL3 was similarly prepared.61 To facilitate the subsequent ring closure metathesis, a Leu-Ser(ψMe,MePro) pseudoproline dipeptide was incorporated in positions 12 and 13. Overall yield of INSL3 analogue was improved and CD spectroscopy of both dicarba isomers was similar to native human INSL3 indicating similar overall secondary structures. Both isomers were equally active in both binding to and activating the INSL3 cognate receptor, RXFP2. The intra-A-chain dicarba human relaxin-2 was also prepared60 and the two isomeric analogues shown to retain near-equipotent RXFP1 receptor binding and activation propensity. Unexpectedly, the in vitro serum stability of both analogues was greatly reduced compared with the native peptide. CD spectroscopy studies showed subtle differences in the secondary structures between the dicarba analogues and human relaxin-2 suggesting that the global fold may be destabilized leading to rapid degradation of dicarba analogues in serum. Caution is therefore recommended when using RCM as a general approach to enhance peptide stability.

Figure 6. Regioselective synthesis of ΔA25/A26 human INSL3 bislactam A11-B10/A24-B22. X = 2,3-diaminopropyl. Conditions: (a) 3% hydrazine/DMF; (b) Fmoc-SPPS; (c) 1% TFA/dichloromethane (DCM); (d) 1.5 equiv of 1-[bis(dimethylamino)methylene]-1H-1,2,3triazolo[4,5-b]pyridinium 3-oxid hexafluorophosphate (HATU)/ N,N-diisopropylethylamine (DIEA); (e) Fmoc-SPPS; (f) TFA cleavage and RP-HPLC; (g) 2 equiv of I2/50% aq. HOAc; (h) RP-HPLC.

Following peptide-resin cleavage and side-chain deprotection, the intra-A-chain disulfide bond (A10−15) was formed by iodine oxidation. The overall yield of purified bis-interchain lactam (ΔA25/26) human INSL3 was low but represented the first documented INSL peptide with lactam crossbridging. In contrast to the triazole-bridged analogue,50 the bis-lactam interchainbridged analogue possessed very strong binding affinity at nanomolar concentration for RXFP2 receptor (pKi 7.92 ± 0.12, cf. the (ΔA25/26) human INSL3s 8.59 ± 0.06) indicating that interchain disulfide bond can be replaced with lactam bridge without altering activity. The tertiary structure and metabolic stability of lactam INSLs remain to be investigated.

Cystathionine-Containing INSLs

Chemical incorporation of redox-stable cystathionine, which is isosterically similar to cystine, into peptides is complex and methods include incorporation of preformed thioether building blocks with orthogonal protection and desulfurization of a disulfide bond. Our approach for replacing the A6−A11 cystine moiety in human insulin with a nonreducible cystathionine thioether bridge63 involved the use of two novel building blocks: a halo-amino acid (Fmoc-γ-bromohomoalanine, Fmoc-γ-BrhAla-OH) and an orthogonally protected cysteine derivative (Nα-monomethoxytrityl-cysteinyl-α-allyl ester, Mmt-Cys-OAll) (Figure 8A,B). After acylation of the halo-amino acid at position 11 within synthetic human insulin A-chain, the thioether moiety was formed via an SN2 reaction with the nucleophilic β-thiol group of Mmt-Cys-Oll. Further peptide synthesis and Nα-deprotection of residue 7 was followed by a lactamization as the cyclization step (Figure 8C) forming a cystathionine bond. Removal of the Nα-Mmt group with 1% TFA in dichloromethane was followed by completion of solid phase synthesis of the

Dicarba-Containing INSLs

Saturated and unsaturated (S,S)-2,7-diaminosuberic residues have been demonstrated to be useful dicarba isosteres of cysteine.54,55 The hydrocarbon cross-links are generated by ruthenium (“Grubb’s”)-catalyzed ring-closing olefin metathesis (RCM) of commercially available allylglycine (Hag).56 This method has proved successful due to the hydrophobic and inert character of those cross-links. RCM is greatly facilitated by microwave irradiation57 and has been adapted to resin-bound peptides following SPPS.58 Upon cleavage of the resin after RCM reaction on solid phase, two isomers typically appear as two peaks on reversed-phase high performance liquid chromatography (RP-HPLC), which correspond to unsaturated cis and trans forms.59−61 On-resin rhodium-catalyzed hydrogenation results 2122

DOI: 10.1021/acs.accounts.7b00288 Acc. Chem. Res. 2017, 50, 2116−2127

Article

Accounts of Chemical Research

Figure 7. (A) Primary structure of intra-A-chain dicarba human relaxin-3. (B) SPPS strategy. (C) RP-HPLC traces of dicarba A-chain (upper) and its resolved isomers (lower). (D) RXFP3 receptor competition (upper) and activation (lower) assays of intra-A-chain dicarba human relaxin-3. Adapted from ref 59. Copyright 2016 Wiley-VCH Verlag GmbH & Co. KGaA.

and lactam.63 Importantly, the thioether analog exhibited enhanced thermal stability when compared to native insulin suggesting that it may be a viable lead analogue for the development of novel diabetes therapeutics for use where refrigeration is limited.

remainder of the A-chain. After a separate assembly of the B-chain and cleavage, deprotection, and purification of both chains, the subsequent formation of the two interchain disulfide bonds followed our established protocols. Overall yield of the intra-A-chain cystathionine insulin analogue was comparable to that for other similar peptides. Critically, native binding affinity for the insulin receptor, secondary structure, and stability in human serum were all maintained indicating that thioether mimics the disulfide bridge in INSLs better than other isosteres including triazole, dicarba,



OUTLOOK The chemical synthesis of INSLs remains a fascinating subject of intense study with the goal of more readily and efficiently acquiring these for chemical biological study as well as for the 2123

DOI: 10.1021/acs.accounts.7b00288 Acc. Chem. Res. 2017, 50, 2116−2127

Article

Accounts of Chemical Research

Figure 8. (A) Synthesis of Fmoc-γ-bromohomoalanine, (B) Nα-monomethoxytrityl-cysteinyl-α-allyl ester, and (C) intra-A-chain cystathionine human insulin. Reagents and conditions: (A) (a) pyridine, (b) HBr/HOAc; (B) (a) DIEA/THF, (b) tri-n-butylphosphine, H2O/THF; (C) (a) Fmoc-γ-BrhAla-OH/HOBt/DIC in DMF, (b) Mmt-Cys-OAll, DIEA/DMF, (c) Fmoc-SPPS, (d) Pd(PPh3)4 in 2.5% N-methylmorpholine (NMM)/5% AcOH/ 92.5% CHCl3, (e) 20% piperidine/DMF, (f) hydroxybenzotriazole (HOBt)/N,N′-diisopropylcarbodiimide (DIC), (g) 1% TFA/DCM, (h) FmocSPPS, (i) 1% triisopropylsilane (TIPS)/2% H2O/2% 3,6-dioxa-1,8-octanedithiol (DODT)/95% TFA, (j) GdHCl, pH 8, (k) I2/AcOH. Abbreviations listed in text.

production of novel analogues. Such developments are essential given that many new INSLs are being identified by genomic

database sequencing and also many modified analogues cannot be readily acquired by recombinant DNA methods. The continued 2124

DOI: 10.1021/acs.accounts.7b00288 Acc. Chem. Res. 2017, 50, 2116−2127

Article

Accounts of Chemical Research

(7) Bathgate, R. A.; Halls, M. L.; van der Westhuizen, E. T.; Callander, G. E.; Kocan, M.; Summers, R. J. Relaxin Family Peptides and their Receptors. Physiol. Rev. 2013, 93, 405−480. (8) Kung, Y. T.; Du, Y. C.; Huang, W. T.; Chen, C. C.; Ke, L. T. Total Synthesis of Crystalline Bovine Insulin. Sci. Sin. 1965, 14, 1710−1716. (9) Du, Y. C.; Jiang, R. Q.; Tsou, C. L. Conditions for Successful Resynthesis of Insulin from its Glycyl and Phenylalanylyl Chains. Sci. Sin. 1965, 14, 229−236. (10) Haugaard-Jonsson, L. M.; Hossain, M. A.; Daly, N. L.; Craik, D. J.; Wade, J. D.; Rosengren, K. J. Structure of Human Insulin-Like Peptide 5 and Characterization of Conserved Hydrogen Bonds and Electrostatic Interactions within the Relaxin Framework. Biochem. J. 2009, 419, 619− 627. (11) Derewenda, U.; Derewenda, Z.; Dodson, G. G.; Hubbard, R. E.; Korber, F. Molecular Structure of Insulin: The Insulin Monomer and its Assembly. Br. Med. Bull. 1989, 45, 4−18. (12) Ward, C. W.; Lawrence, M. C. Landmarks in Insulin Research. Front. Endocrinol. 2011, 2, 76. (13) Chang, S. G.; Choi, K. D.; Jang, S. H.; Shin, H. C. Role of Disulfide Bonds in the Structure and Activity of Human Insulin. Mol. Cells 2003, 16, 323−330. (14) Patil, N. A.; Tailhades, J.; Hughes, R. A.; Separovic, F.; Wade, J. D.; Hossain, M. A. Cellular Disulfide Bond Formation in Bioactive Peptides and Proteins. Int. J. Mol. Sci. 2015, 16, 1791−1805. (15) Marglin, A.; Merrifield, R. B. The Synthesis of Bovine Insulin by the Solid Phase Method. J. Am. Chem. Soc. 1966, 88, 5051−5052. (16) Tregear, G.; Du, Y. C.; Fagan, C.; Reynolds, D.; Scanlon, D.; Jones, P.; Kemp, B.; Niall, H. Porcine Relaxin: Synthesis and Structure Activity Relationships. In Peptides; Rich, D. H., Gross, E., Eds.; Pierce: Rockford, 1981; pp 249−252. (17) Hudson, P.; John, M.; Crawford, R.; Haralambidis, J.; Scanlon, D.; Gorman, J.; Tregear, G.; Shine, J.; Niall, H. Relaxin Gene Expression in Human Ovaries and the Predicted Structure of a Human Preprorelaxin by Analysis of cDNA Clones. EMBO J. 1984, 3, 2333−2339. (18) Sieber, P.; Kamber, B.; Hartmann, A.; Johl, A.; Riniker, B.; Rittel, W. [Total Synthesis of Human Insulin. IV. Description of the Final Steps (author’s transl.)]. Helv. Chim. Acta 1977, 60, 27−37. (19) Maruyama, K.; Nagata, K.; Tanaka, M.; Nagasawa, H.; Isogai, A.; Ishizaki, H.; Suzuki, A. Synthesis of Bombyxin-IV, an Insulin Superfamily Peptide from the Silkworm, Bombyx mori, by Stepwise and Selective Formation of Three Disulfide Bridges. J. Protein Chem. 1992, 11, 1−12. (20) Akaji, K.; Fujino, K.; Tatsumi, T.; Kiso, Y. Total Synthesis of Human Insulin by Regioselective Disulfide Formation using the Silyl Chloride-Sulfoxide Method. J. Am. Chem. Soc. 1993, 115, 11384−11392. (21) Büllesbach, E. E.; Schwabe, C. Total Synthesis of Human Relaxin and Human Relaxin derivatives by Solid-Phase Peptide Synthesis and Site-Directed Chain Combination. J. Biol. Chem. 1991, 266, 10754− 10761. (22) Bathgate, R. A. D.; Lin, F.; Hanson, N. F.; Otvos, L.; Guidolin, A.; Giannakis, C.; Bastiras, S.; Layfield, S. L.; Ferraro, T.; Ma, S.; Zhao, C. X.; Gundlach, A. L.; Samuel, C. S.; Tregear, G. W.; Wade, J. D. Relaxin-3: Improved Synthesis Strategy and Demonstration of its High-Affinity Interaction with the Relaxin Receptor LGR7 Both In Vitro and In Vivo. Biochemistry 2006, 45, 1043−1053. (23) Rosengren, K. J.; Lin, F.; Bathgate, R. A.; Tregear, G. W.; Daly, N. L.; Wade, J. D.; Craik, D. J. Solution Structure and Novel Insights into the Determinants of the Receptor Specificity of Human Relaxin-3. J. Biol. Chem. 2006, 281, 5845−5851. (24) Rosengren, K. J.; Zhang, S.; Lin, F.; Daly, N. L.; Scott, D. J.; Hughes, R. A.; Bathgate, R. A.; Craik, D. J.; Wade, J. D. Solution Structure and Characterization of the LGR8 Receptor Binding Surface of Insulin-Like Peptide 3. J. Biol. Chem. 2006, 281, 28287−28295. (25) Haugaard-Jonsson, L. M.; Hossain, M. A.; Daly, N. L.; Bathgate, R. A.; Wade, J. D.; Craik, D. J.; Rosengren, K. J. Structure of the R3/I5 Chimeric Relaxin Reptide, a Selective GPCR135 and GPCR142 Agonist. J. Biol. Chem. 2008, 283, 23811−23818. (26) Haugaard-Kedstrom, L. M.; Hossain, M. A.; Daly, N. L.; Bathgate, R. A.; Rinderknecht, E.; Wade, J. D.; Craik, D. J.; Rosengren, K. J.

development of new cysteine S-thiol protecting groups with novel orthogonality together with improved solid phase synthesis and regioselective disulfide bond formation augur well for the future improved preparation of these peptides including those with two and three disulfide bond isosteres that confer increased in vivo stability.



AUTHOR INFORMATION

Corresponding Authors

*E-mail: akhter.hossain@florey.edu.au. *E-mail: john.wade@florey.edu.au. ORCID

John D. Wade: 0000-0002-1352-6568 Funding

Part of the reported studies was funded by the NHMRC of Australia through Project grants to M.A.H. and J.D.W. and a Principal Research Fellowship to JDW (APP628404 and 1117483). Research at The Florey Institute of Neuroscience and Mental Health is supported by the Victorian Government Operational Infrastructure Support Program Notes

The authors declare no competing financial interest. Biographies Mohammed Akhter Hossain received his Ph.D. degree in 2001 from the Tokyo Institute of Technology on chemosensors based on cyclodextrin−peptide conjugates with Prof Akihiko Ueno. He was then a postdoctoral fellow at the Josef Fourier University in France. In 2005, he joined the Florey Institute of Neuroscience and Mental Health in Melbourne, Australia, where his research interests involve the design and synthesis of novel insulin and relaxin peptidomimetics, and β-amyloid peptides. He is currently a Florey Senior Research Fellow and head of the Insulin Peptides Laboratory. John D. Wade obtained his Ph.D. in 1979 at Monash University, Australia, on the structural basis of the diabetogenic action of growth hormone. He received a Nuffield Foundation Fellowship to Cambridge, U.K., to the laboratory of Dr. R. C. Sheppard at the MRC Laboratory of Molecular Biology on the development of the Fmoc-SPPS methodology. In 1983, he returned to Australia to the Florey Institute of Neuroscience and Mental Health, University of Melbourne, where he heads the Laboratory of Peptide Chemistry. His interests include SPPS chemistry of large, complex peptides. Professor Wade is an NHMRC of Australia Principal Research Fellow and a Fellow of the Royal Australian Chemical Institute and of the Royal Society of Chemistry.



REFERENCES

(1) Ryle, A. P.; Sanger, F.; Smith, L. F.; Kitai, R. The Disulphide Bonds of Insulin. Biochem. J. 1955, 60, 541−556. (2) Schwabe, C.; McDonald, J. K. Relaxin: a Disulfide Homolog of Insulin. Science 1977, 197, 914−915. (3) Shabanpoor, F.; Separovic, F.; Wade, J. D. The Human Insulin Superfamily of Polypeptide Hormones. Vitam. Horm. 2009, 80, 1−31. (4) Claeys, I.; Simonet, G.; Poels, J.; Van Loy, T.; Vercammen, L.; De Loof, A.; Vanden Broeck, J. Insulin-Related Peptides and their Conserved Signal Transduction Pathway. Peptides 2002, 23, 807−816. (5) Pinero-Gonzalez, J.; Gonzalez-Perez, A. The Ubiquity of the Insulin Superfamily across the Eukaryotes Detected using a Bioinformatics Approach. OMICS 2011, 15, 439−447. (6) Patil, N. A.; Rosengren, K. J.; Separovic, F.; Wade, J. D.; Bathgate, R. A.; Hossain, M. A. Relaxin Family Peptides: Structure-Activity Relationship Studies. Br. J. Pharmacol. 2017, 174, 950−961. 2125

DOI: 10.1021/acs.accounts.7b00288 Acc. Chem. Res. 2017, 50, 2116−2127

Article

Accounts of Chemical Research Solution Structure, Aggregation Behavior, and Flexibility of Human Relaxin-2. ACS Chem. Biol. 2015, 10, 891−900. (27) Hossain, M. A.; Lin, F.; Zhang, S. D.; Ferraro, T.; Bathgate, R. A.; Tregear, G. W.; Wade, J. D. Regioselective Disulfide Solid Phase Synthesis, Chemical Characterization and In Vitro Receptor Binding Activity of Equine Relaxin. Int. J. Pept. Res. Ther. 2006, 12, 211−215. (28) Lin, F.; Tailhades, J.; Chan, L. J.; Bathgate, R. A. D.; Hossain, M. A.; Wade, J. D. Preparation of Canine Relaxin by Fmoc-Solid Phase Synthesis and Regioselective Disulfide Bond Formation within the Aand B-Chains. Biochem. Comp. 2013, 1, 4. (29) Samuel, C. S.; Lin, F.; Hossain, M. A.; Zhao, C.; Ferraro, T.; Bathgate, R. A.; Tregear, G. W.; Wade, J. D. Improved Chemical Synthesis and Demonstration of the Relaxin Receptor Binding Affinity and Biological Activity of Mouse Relaxin. Biochemistry 2007, 46, 5374− 5381. (30) Shabanpoor, F.; Hossain, M. A.; Ryan, P. J.; Belgi, A.; Layfield, S.; Kocan, M.; Zhang, S.; Samuel, C. S.; Gundlach, A. L.; Bathgate, R. A.; Separovic, F.; Wade, J. D. Minimization of Human Relaxin-3 Leading to High-Affinity Analogues with Increased Selectivity for Relaxin-Family Peptide 3 Receptor (RXFP3) over RXFP1. J. Med. Chem. 2012, 55, 1671−1681. (31) Hossain, M. A.; Bathgate, R. A. D.; Rosengren, K. J.; Shabanpoor, F.; Zhang, S. D.; Lin, F.; Tregear, G. W.; Wade, J. D. The Structural and Functional Role of the B-chain C-terminal Arginine in the Relaxin-3 Peptide Antagonist, R3(B Delta 23−27)R/I5. Chem. Biol. Drug Des. 2009, 73, 46−52. (32) Hossain, M. A.; Rosengren, K. J.; Samuel, C. S.; Shabanpoor, F.; Chan, L. J.; Bathgate, R. A.; Wade, J. D. The Minimal Active Structure of Human Relaxin-2. J. Biol. Chem. 2011, 286, 37555−37565. (33) Shabanpoor, F.; Hughes, R. A.; Bathgate, R. A. D.; Zhang, S.; Scanlon, D. B.; Lin, F.; Hossain, M. A.; Separovic, F.; Wade, J. D. SolidPhase Synthesis of Europium-Labeled Human INSL3 as a Novel Probe for the Study of Ligand-Receptor Interactions. Bioconjugate Chem. 2008, 19, 1456−1463. (34) Chan, L. D. J.; Smith, C. M.; Chua, B. E.; Lin, F.; Bathgate, R. A. D.; Separovic, F.; Gundlach, A. L.; Hossain, M. A.; Wade, J. D. Synthesis of Fluorescent Analogs of Relaxin Family Peptides and their Preliminary In Vitro and In Vivo Characterization. Front. Chem. 2013, 1, 30. (35) Shabanpoor, F.; Bathgate, R. A.; Belgi, A.; Chan, L. J.; Nair, V. B.; Wade, J. D.; Hossain, M. A. Site-Specific Conjugation of a Lanthanide Chelator and its Effects on the Chemical Synthesis and Receptor Binding Affinity of Human Relaxin-2 hormone. Biochem. Biophys. Res. Commun. 2012, 420, 253−256. (36) Lin, F.; Hossain, M. A.; Post, S.; Karashchuk, G.; Tatar, M.; De Meyts, P.; Wade, J. D. Total Solid-Phase Synthesis of Biologically Active Drosophila Insulin-Like Peptide 2 (DILP2). Aust. J. Chem. 2017, 70, 208−212. (37) Hossain, M. A.; Belgi, A.; Lin, F.; Zhang, S.; Shabanpoor, F.; Chan, L.; Belyea, C.; Truong, H. T.; Blair, A. R.; Andrikopoulos, S.; Tregear, G. W.; Wade, J. D. Use of a Temporary ″Solubilizing″ Peptide Tag for the Fmoc Solid-Phase Synthesis of Human Insulin Glargine via use of Regioselective Disulfide Bond Formation. Bioconjugate Chem. 2009, 20, 1390−1396. (38) Adams, S. R.; Tsien, R. Y. Controlling Cell Chemistry with Caged Compounds. Annu. Rev. Physiol. 1993, 55, 755−784. (39) Klan, P.; Solomek, T.; Bochet, C. G.; Blanc, A.; Givens, R.; Rubina, M.; Popik, V.; Kostikov, A.; Wirz, J. Photoremovable Protecting Groups in Chemistry and Biology: Reaction Mechanisms and Efficacy. Chem. Rev. 2013, 113, 119−191. (40) Karas, J. A.; Scanlon, D. B.; Forbes, B. E.; Vetter, I.; Lewis, R. J.; Gardiner, J.; Separovic, F.; Wade, J. D.; Hossain, M. A. 2-Nitroveratryl as a Photocleavable Thiol-Protecting Group for Directed Disulfide Bond Formation in the Chemical Synthesis of Insulin. Chem. - Eur. J. 2014, 20, 9549−9552. (41) Liu, F.; Luo, E. Y.; Flora, D. B.; Mezo, A. R. A Synthetic Route to Human Insulin Using Isoacyl Peptides. Angew. Chem., Int. Ed. 2014, 53, 3983−3987.

(42) Kotzur, N.; Briand, B.; Beyermann, M.; Hagen, V. WavelengthSelective Photoactivatable Protecting Groups for Thiols. J. Am. Chem. Soc. 2009, 131, 16927−16931. (43) Sohma, Y.; Hua, Q. X.; Whittaker, J.; Weiss, M. A.; Kent, S. B. H. Design and Folding of [Glu(A4)(O(beta)Thr(B30))]Insulin (″Ester Insulin″): A Minimal Proinsulin Surrogate that Can Be Chemically Converted into Human Insulin. Angew. Chem., Int. Ed. 2010, 49, 5489− 5493. (44) Zaykov, A. N.; Mayer, J. P.; Gelfanov, V. M.; DiMarchi, R. D. Chemical Synthesis of Insulin Analogs through a Novel Precursor. ACS Chem. Biol. 2014, 9, 683−691. (45) Thalluri, K.; Kou, B.; Gelfanov, V.; Mayer, J. P.; Liu, F.; DiMarchi, R. D. Biomimetic Synthesis of Insulin Enabled by Oxime Ligation and Traceless ″C-Peptide″ Chemical Excision. Org. Lett. 2017, 19, 706−709. (46) Zaykov, A. N.; Mayer, J. P.; DiMarchi, R. D. Pursuit of a Perfect Insulin. Nat. Rev. Drug Discovery 2016, 15, 425−439. (47) Patil, N. A.; Tailhades, J.; Karas, J. A.; Separovic, F.; Wade, J. D.; Hossain, M. A. A One-Pot Chemically Cleavable Bis-Linker Tether Strategy for the Synthesis of Heterodimeric Peptides. Angew. Chem., Int. Ed. 2016, 55, 14552−14556. (48) Muttenthaler, M.; Andersson, A.; de Araujo, A. D.; Dekan, Z.; Lewis, R. J.; Alewood, P. F. Modulating Oxytocin Activity and Plasma Stability by Disulfide Bond Engineering. J. Med. Chem. 2010, 53, 8585− 8596. (49) Holland-Nell, K.; Meldal, M. Maintaining Biological Activity by Using Triazoles as Disufide Bond Mimetics. Angew. Chem., Int. Ed. 2011, 50, 5204−5206. (50) Williams, G. M.; Lee, K.; Li, X.; Cooper, G. J.; Brimble, M. A. Replacement of the CysA7-CysB7 Disulfide Bond with a 1,2,3-Triazole Linker Causes Unfolding in Insulin Glargine. Org. Biomol. Chem. 2015, 13, 4059−4063. (51) Dong, M.; Te, J. A.; Xu, X.; Wang, J.; Pinon, D. I.; Storjohann, L.; Bordner, A. J.; Miller, L. J. Lactam Constraints Provide Insights into the Receptor-Bound Conformation of Secretin and Stabilize a Receptor Antagonist. Biochemistry 2011, 50, 8181−8192. (52) Lanigan, M. D.; Pennington, M. W.; Lefievre, Y.; Rauer, H.; Norton, R. S. Designed Peptide Analogues of the Potassium Channel Blocker ShK Toxin. Biochemistry 2001, 40, 15528−15537. (53) Karas, J.; Shabanpoor, F.; Hossain, M. A.; Gardiner, J.; Separovic, F.; Wade, J. D.; Scanlon, D. B. Total Chemical Synthesis of a Heterodimeric Interchain Bis-Lactam-linked Peptide: Application to an Analogue of Human Insulin-Like Peptide 3. Int. J. Pept. 2013, 2013, 504260. (54) Brik, A. Metathesis in Peptides and Peptidomimetics. Adv. Synth. Catal. 2008, 350, 1661−1675. (55) Li, P.; Roller, P. P.; Xu, J. Current Synthetic Approaches to Peptide and Peptidomimetic Cyclization. Curr. Org. Chem. 2002, 6, 411−440. (56) Robinson, A. J.; van Lierop, B. J.; Garland, R. D.; Teoh, E.; Elaridi, J.; Illesinghe, J. P.; Jackson, W. R. Regioselective Formation of Interlocked Dicarba Bridges in Naturally Occurring Cyclic Peptide Toxins using Olefin Metathesis. Chem. Commun. 2009, 4293−4295. (57) Robinson, A. J.; Elaridi, J.; Van Lierop, B. J.; Mujcinovic, S.; Jackson, W. R. Microwave-Assisted RCM for the Synthesis of Carbocyclic Peptides. J. Pept. Sci. 2007, 13, 280−285. (58) Stymiest, J. L.; Mitchell, B. F.; Wong, S.; Vederas, J. C. Synthesis of Biologically Active Dicarba Analogues of the Peptide Hormone Oxytocin using Ring-Closing Metathesis. Org. Lett. 2003, 5, 47−49. (59) Hossain, M. A.; Rosengren, K. J.; Zhang, S.; Bathgate, R. A.; Tregear, G. W.; van Lierop, B. J.; Robinson, A. J.; Wade, J. D. Solid Phase Synthesis and Structural Analysis of Novel A-Chain Dicarba Analogs of Human Relaxin-3 (INSL7) that Exhibit Full Biological Activity. Org. Biomol. Chem. 2009, 7, 1547−1553. (60) Hossain, M. A.; Haugaard-Kedstrom, L. M.; Rosengren, K. J.; Bathgate, R. A.; Wade, J. D. Chemically Synthesized Dicarba H2 Relaxin Analogues Retain Strong RXFP1 Receptor Activity but Show an Unexpected Loss of In Vitro Serum Stability. Org. Biomol. Chem. 2015, 13, 10895−10903. 2126

DOI: 10.1021/acs.accounts.7b00288 Acc. Chem. Res. 2017, 50, 2116−2127

Article

Accounts of Chemical Research (61) Zhang, S.; Hughes, R. A.; Bathgate, R. A.; Shabanpoor, F.; Hossain, M. A.; Lin, F.; van Lierop, B.; Robinson, A. J.; Wade, J. D. Role of the Intra-A-Chain Disulfide Bond of Insulin-Like Peptide 3 in Binding and Activation of its Receptor, RXFP2. Peptides 2010, 31, 1730−1736. (62) Cui, H.-K.; Guo, Y.; He, Y.; Wang, F.-L.; Chang, H.-N.; Wang, Y.J.; Wu, F.-M.; Tian, C.-L.; Liu, L. Diaminodiacid-Based Solid-Phase Synthesis of Peptide Disulfide Bond Mimics. Angew. Chem., Int. Ed. 2013, 52, 9558−9562. (63) Karas, J. A.; Patil, N. A.; Tailhades, J.; Sani, M. A.; Scanlon, D. B.; Forbes, B. E.; Gardiner, J.; Separovic, F.; Wade, J. D.; Hossain, M. A. Total Chemical Synthesis of an Intra-A-Chain Cystathionine Human Insulin Analogue with Enhanced Thermal Stability. Angew. Chem., Int. Ed. 2016, 55, 14743−14747.

2127

DOI: 10.1021/acs.accounts.7b00288 Acc. Chem. Res. 2017, 50, 2116−2127