Flexible Folding: Disulfide-Containing Peptides and Proteins Choose The

molecules

Review Flexible Folding: Disulﬁde-Containing Peptides and Proteins Choose the Pathway Depending on the Environments

Kenta Arai and Michio Iwaoka *

Department of Chemistry, School of Science, Tokai University, Kitakaname, Hiratsuka-shi, Kanagawa 259-1292, Japan; [email protected] * Correspondence: [email protected]; Tel.: +81-463-58-1211

Abstract: In the last few decades, development of novel experimental techniques, such as new types of disulfide (SS)-forming reagents and genetic and chemical technologies for synthesizing designed artificial proteins, is opening a new realm of the oxidative folding study where peptides and proteins can be folded under physiologically more relevant conditions. In this review, after a brief overview of the historical and physicochemical background of oxidative protein folding study, recently revealed folding pathways of several representative peptides and proteins are summarized, including those having two, three, or four SS bonds in the native state, as well as those with odd Cys residues or consisting of two peptide chains. Comparison of the updated pathways with those reported in the early years has revealed the flexible nature of the protein folding pathways. The significantly different pathways characterized for hen-egg white lysozyme and bovine milk α-lactalbumin, which belong to the same protein superfamily, suggest that the information of protein folding pathways, not only the native folded structure, is encoded in the amino acid sequence. The application of the flexible pathways of peptides and proteins to the engineering of folded three-dimensional structures

is an interesting and important issue in the new realm of the current oxidative protein folding study. Keywords: protein folding; neurodegenerative diseases; protein engineering; selenoxide; selenopeptides Citation: Arai, K.; Iwaoka, M. Flexible Folding: Disulfide- Containing Peptides and Proteins Choose the Pathway Depending on the Environments. Molecules 2021, 26, 1. Introduction 195. https://doi.org/10.3390/ Proteins are synthesized in cells as sequential strings of amino acids connected via molecules26010195 peptide bonds. However, the linear peptide chains are in most cases inactive in terms of the biological functions because proteins exert their biological activities only when Academic Editor: Derek J. McPhee they are converted into the unique three-dimensional (3D) structures. This structure- Received: 12 December 2020 forming process of a polypeptide chain is called protein folding, and is known to proceed Accepted: 29 December 2020 spontaneously in a test tube for many small globular proteins. Published: 2 January 2021 Since the pioneering studies by Anfinsen [1,2] and Levinthal [3] in the 1960s, researchers have exerted enormous efforts to elucidate the folding pathways of various Publisher’s Note: MDPI stays neu- proteins. However, it is still not easy to characterize the intermediates, especially when a tral with regard to jurisdictional clai- protein is folded with its disulfide (SS) bonds intact due to the short time constant of the ms in published maps and institutio- folding reaction (<1 s). On the other hand, an oxidative folding from the SS-reduced state nal affiliations. takes place with a much longer time constant (>1 s). Thus, the SS-coupled protein folding is advantageous in detecting the folding intermediates, with which we can delineate a scenario of protein folding. Copyright: © 2021 by the authors. Li- In early studies of SS-coupled (or oxidative) protein folding, researchers focused on censee MDPI, Basel, Switzerland. the physicochemical aspects, i.e., observation and characterization of the folding interme- This article is an open access article diates, although the reactions were frequently monitored under the conditions slightly distributed under the terms and con- different in terms of the temperature and pH from those in vivo due to the technical limita- ditions of the Creative Commons At- tions [4]. On the other hand, thanks to the novel experimental techniques developed in tribution (CC BY) license (https:// the last few decades, the oxidative folding study is moving to a new realm. In short, new creativecommons.org/licenses/by/ types of oxidants (SS-forming reagents), as well as genetic and chemical technologies for 4.0/).

Molecules 2021, 26, 195. https://doi.org/10.3390/molecules26010195 https://www.mdpi.com/journal/molecules MoleculesMoleculesMolecules 2021 20212021,,, , 26 2626,,, , x xxx FOR FORFORFOR PEER PEERPEERPEER REVIEW REVIEWREVIEWREVIEW 222 of of of 20 20 20 Molecules Molecules 20212021,, 2626,, xx FORFOR PEERPEER REVIEWREVIEW 22 of of 20 20 Molecules 2021, 26, x FOR PEER REVIEW 2 of 20 Molecules 2021,, 26,, xx FORFOR PEERPEER REVIEWREVIEW 2 of 20

differentdifferentdifferent in inin terms termsterms of ofof the thethe temperature temperaturetemperature and andand pH pHpH from fromfrom those thosethose in ininin vivo vivovivovivo due dueduedue to tototo the thethethe technical technicaltechnicaltechnical limi- limi-limi-limi- differentdifferent inin termsterms ofof thethe temperaturetemperature andand pHpH fromfrom thosethose inin vivovivo duedue toto thethe technicaltechnical limi-limi- tationsdifferenttationstationsdifferenttations [4]. [4].[4].[4]. inin On OnOnOn termsterms the thethethe of other ofotherotherother thethe hand,temperature hand,hand,temperaturehand, thanks thanksthanksthanks and andto tototo th thth th pHpHeee novel novelnovel fromfrom experimentalexperimental thoseexperimentalthose inin vivovivo due techniques duetechniquestechniques toto thethe technical technical developed developeddeveloped limi-limi- in inin Molecules 2021, 26, 195 differenttationstations [4].[4]. in OnOnterms thethe of otherother the temperature hand,hand, thanksthanks and toto th thpHee novel novelfrom experimentalthoseexperimental in vivo due techniquestechniques to the technical developeddeveloped2 of 19limi- inin thetationsthethetationsthe last lastlastlast [4]. few [4].fewfewfew On On decades, decades,decades,decades, thethe otherother the thethethe hand, oxidative hand,oxidativeoxidativeoxidative thanksthanks folding foldingfoldingfolding toto thth study estudystudyestudy novelnovel is isisis experimental experimentalmoving movingmovingmoving to tototo a aaa new newnew newtechniquestechniques realm. realm.realm.realm. In IndevelopedInIndeveloped short, short,short,short, new newnewnew inin tationsthethe lastlast [4]. fewfew On decades,decades, the other thethe hand, oxidativeoxidative thanks foldingfolding to th e studystudy novel isis experimental movingmoving toto aa newtechniquesnew realm.realm. developedInIn short,short, newnew in typesthetypestypesthetypes lastlast of ofofof few few oxidants oxidantsoxidantsoxidants decades,decades, (SS-forming (SS-forming(SS-forming(SS-forming thethe oxidativeoxidative reagents), reagents),reagents),reagents), foldingfolding as asasas study studywe wewewellllllll as asas asisis genetic movinggeneticgeneticgeneticmoving and andandand toto aa chemicalchemicalchemical newchemicalnew realm.realm. technologies technologiestechnologiestechnologies InIn short,short, newnew for forforfor thetypestypes last ofof few oxidantsoxidants decades, (SS-forming(SS-forming the oxidative reagents),reagents), folding asas study wewellll asisas movinggeneticgenetic andandto a chemicalnewchemical realm. technologiestechnologies In short, new forfor synthesizingtypessynthesizingtypessynthesizing ofof oxidantsoxidants designed designeddesigned (SS-forming(SS-forming artifici artificiartificialalal reagents), reagents), proteins, proteins,proteins, have ashavehaveas wewe been beenllbeenll asas geneticdevelo develogeneticdeveloped.ped. ped. andand TheThe chemicalThechemical new newnew oxidants oxidants oxidants technologiestechnologies enabled enabledenabled forfor synthesizingtypessynthesizingsynthesizing of oxidants designed designeddesigned (SS-forming artificial artificiartifici proteins,al alreagents), proteins,proteins, have as havehave been we beenllbeen developed.as genetic develodeveloped. ped.and The ThechemicalThe new newnew oxidants oxidants oxidantstechnologies enabled enabledenabled for thesynthesizingthethesynthesizingthe SS-coupledSS-coupledSS-coupledSS-coupled designed designed foldingfoldingfoldingfolding artifici artifici totototo bebebebe performedalperformedperformedalperformed proteins, proteins, underhave underunderunderhave been been biologicallybiologicallybiologicallybiologically develo developed.ped. moremoremoremore The The relevantrelevantrelevantrelevant new new oxidants oxidants conditions.conditions.conditions.conditions. enabled enabled ByByByBy thesynthesizingthethe SS-coupled SS-coupledSS-coupled foldingdesigned foldingfolding to artifici be toto performed bebe al performedperformed proteins, under haveunderunder biologically been biologicallybiologically develo moreped. moremore relevantThe relevantrelevant new conditions. oxidants conditions.conditions. enabled By ByBy applyingtheapplyingtheapplying SS-coupledSS-coupled suchsuchsuch newnew foldingnewfolding tetetechniques,chniques,chniques, toto bebe performedperformed unprecedentedunprecedentedunprecedented underunder biologically biologically foldingfoldingfolding pathwayspathwayspathways moremore relevant relevant havehavehave beenbeenbeen conditions.conditions. discovereddiscovereddiscovered ByBy applyingtheapplyingapplying SS-coupled such suchsuch new newfoldingnew techniques, tetechniques,chniques, to be unprecedentedperformed unprecedentedunprecedented under folding biologically foldingfolding pathways pathwayspathways more have relevant havehave been beenbeen discoveredconditions. discovereddiscovered By forforapplyingforfor severalapplyingfor several severalseveralseveral SS-containing suchsuch SS-containing SS-containingSS-containingSS-containing newnew tetechniques, proteins,chniques, protei proteiproteiproteins,ns, suggestingns, unprecedentedunprecedented suggesting suggestingsuggesting that that thatthat folding foldingfolding folding foldingfolding pathways pathwayspathways pathways pathwayspathways of havehave proteins of ofof proteins proteinsproteins beenbeen are discovereddiscovered are are muchare much muchmuch applyingforfor severalseveral such SS-containingSS-containing new techniques, proteiproteins, ns,unprecedented suggestingsuggesting thatthat folding foldingfolding pathways pathwayspathways have ofof proteinsproteins been discovered areare muchmuch moremoreformoreformoreflexible several several flexible flexibleflexible thanSS-containing SS-containing than thanthan those those thosethose the the the researcherstheproteiprotei researchers researchersresearchersns,ns, suggesting suggesting proposed proposed proposedproposed that that previously. previously. foldingpreviously.previously.folding pathways pathways of of proteins proteins are are much much formoremore several flexibleflexible SS-containing thanthan thosethose theproteithe researchersresearchersns, suggesting proposedproposed that folding previously.previously. pathways of proteins are much moremoreIn thisInInIn In flexibleflexible this thisthisthis review, review, review,review,review, thanthan after those those after afterafterafter a brief thea aathea brief briefbriefbrief researchers overviewresearchers overview overviewoverviewoverview of proposedproposed the of ofofof the thethe historicalthe historical historicalhistoricalhistorical previously.previously. and and andandand physicochemical physicochemical physicochemicalphysicochemicalphysicochemical aspects aspects aspectsaspectsaspects of of ofofof moreIn Inflexible thisthis review,review, than those afterafter the aa briefbrief researchers overviewoverview proposed ofof thethe historicalhistorical previously. andand physicochemicalphysicochemical aspectsaspects ofof oxidativeoxidativeoxidativeoxidativeInIn protein thisthis protein proteinprotein review,review, folding folding foldingfolding afterafter study, a study,a study,study, briefbrief the overviewthe theoverview recentlythe recently recentlyrecently revised ofof revised revisedtherevisedthe historicalhistorical oxidative oxidative oxidativeoxidative andand folding folding folding physicochemicalfoldingphysicochemical pathways pathways pathwayspathways of several aspectsof ofaspectsof several severalseveral ofof oxidativeoxidativeIn this proteinprotein review, foldingfolding after a study,study, brief overview thethe recentlyrecently of the revisedrevised historical oxidativeoxidative and foldingphysicochemicalfolding pathwayspathways aspects ofof severalseveral of representativerepresentativeoxidativerepresentativeoxidativerepresentative protein protein peptides peptidespeptidespeptides folding folding and andand andstudy, study, proteins proteinsproteinsproteins the the (Tablerecently recently (Table(Table(Table1) arerevised revised 1)1)1) arecomparedareare comparedcomparedoxidative comparedoxidative with folding folding withwithwith those thosethosethose pathways pathways reported reportedreportedreported inof of several the several ininin thethethe oxidativerepresentativerepresentative protein peptidespeptides folding and andstudy, proteinsproteins the recently (Table(Table revised 1)1) areare comparedoxidativecompared folding withwith thosethose pathways reportedreported of several inin thethe earlyearlyrepresentativeearlyrepresentativeearly years, years,years,years, focusing focusingfocusingfocusing peptidespeptides on onon theon and andthethethe flexible flexibleproteinsflexibleproteinsflexible nature nature naturenature (Table(Table of protein ofof 1)of1) protein proteinareproteinare compared foldingcompared foldingfoldingfolding pathways. with with pathways.pathways.pathways. thosethose Some reported reported SomeSomeSome possible possiblepossiblepossible inin thethe representativeearlyearly years,years, focusingfocusing peptides onon and thethe proteinsflexibleflexible naturenature(Table of1)of proteinareprotein compared foldingfolding with pathways.pathways. those reported SomeSome possiblepossible in the applicationsapplicationsapplicationsearlyapplications years, of the offocusingofof thethethe flexible flexibleflexibleflexible on folding the foldingfoldingfolding flexible to engineering tototo nature engineeringengineeringengineering of theprotein thethethe peptide peptidepeptidepeptide folding and andand andpathways. protein proteinproteinprotein structures Some structuresstructuresstructures possible are areareare earlyapplicationsapplications years, focusing ofof thethe flexibleflexible on the foldingfolding flexible toto nature engineeringengineering of protein thethe peptidepeptidefolding andpathways.and proteinprotein Some structuresstructures possible areare proposedapplicationsproposedapplications as as intriguing intriguingofof thethe flexibleflexible issues issues foldingfolding in in the the totonew new engineeringengineering realm realm of of oxidative oxidative thethe peptidepeptide protein protein andand foldingprotein proteinfolding study. structuresstructuresstudy. areare proposedproposedapplicationsproposed as intriguing asas intriguingofintriguing the flexible issues issuesissues infolding the inin thethe new to newnew realmengineering realmrealm of oxidative ofof oxidativeoxidativethe peptide protein proteinprotein and folding protein foldingfolding study. structures study.study. are proposedproposed asas intriguingintriguing issuesissues inin thethe newnew realmrealm ofof oxidativeoxidative proteinprotein foldingfolding study.study. proposed as intriguing issues in the new realm of oxidative protein folding study. TableTableTableTable 1. Model 1. 1.1. Model ModelModel peptides peptides peptidespeptides and and andand proteins proteins proteinsproteins for thefofoforrr the oxidativethethe oxidative oxidativeoxidative folding folding foldingfolding study. study. study.study. TableTable 1.1. ModelModel peptidespeptides andand proteinsproteins foforr thethe oxidativeoxidative foldingfolding study.study. TableTable 1.1. ModelModel peptidespeptides andand proteins proteins foforr thethe oxidativeoxidative foldingfolding study.study. TableNumber 1. ModelNumberNumberNumber of SSpeptides of ofof andNumberNumberNumberNumber proteins of Cys of ofoffo r the oxidative folding study. CategoryCategoryCategory ProteinsProteinsProteins NumberNumber of of NumberNumber of of SS-BondSS-Bond TopologyTopology Topology Section Section Section Category Proteins BondsNumberNumberSSSS bonds bonds ofof CysCysNumberNumberResidues Residues Residues ofof SS-Bond Topology Section CategoryCategory ProteinsProteins NumberSSSS bondsbonds of CysCysNumber ResiduesResidues of SS-BondSS-Bond TopologyTopology Section Section Category Proteins SSSS bondsbonds CysCys ResiduesResidues H-H-SS-Bond Topology -OH -OH Section Category Proteins SS bonds Cys Residues H-H-SS-Bond Topology -OH -OH Section insulininsulin A-chaininsulin A-chain A-chain 2SS bonds 2 2 Cys Residues 4 4 4 H-H-C6C6 C7 C7 C11C11 C20C20 -OH -OH insulininsulininsulin A-chainA-chainA-chain 2 2 2 4 4 4 H-C6C6 C7 C7 C11C11 C20C20 -OH H-C20 -OH insulininsulin A-chainA-chain 2 2 4 4 H-C6C6 C7 C7 C11C11 C20 -OH insulin A-chain 2 4 H-H-C6 C7 C11 C20 -OH -OH insulin A-chain 2 4 H-H-C6 C7 C11 C20 -OH -OH relaxinrelaxin A-chainrelaxin A-chain A-chain 2 2 2 4 4 4 H-H-C10C10C11C11 C16C16 C24C24 -OH -OH relaxin A-chain 2 4 H-H-C10 C11 C16 C24 -OH -OH 2SS 2SS2SS2SS relaxinrelaxin A-chainA-chain 2 2 4 4 C10C10 C11C11 C16C16 C24C24 4.1 4.14.14.1 2SS relaxin A-chain 2 4 H-SSC10SS C11 C16 C24 -OH 4.1 2SS2SS relaxin A-chain 2 4 SSSSSS 4.14.1 2SS relaxin A-chain 2 4 C10SSSSC11 C16 C24 4.1 2SS relaxin A-chain 2 4 SSSSSSSSSS 4.1 2SS SSSS 4.1 endothelin-1endothelin-1 22 4 4 H-H- SS SSSS -OH-OH endothelin-1endothelin-1endothelin-1 222 4 4 4 H- SS SS -OH-OH endothelin-1 2 4 H-H-C1C1 C3 C3 C11 C11 C15C15 -OH-OH endothelin-1endothelin-1 22 4 4 H-H-C1C1 C3 C3SS C11 C11 C15C15 -OH-OH endothelin-1 2 4 H- C1C1 C3 C3 C11 C11 C15C15 -OH endothelin-1 2 4 H- SS -OH C1 C3SSSSSS C11SS C15 C1 C3SS C11SSSSSS C15 C1 C3SSSS C11SS C15 SSSS SS SS SS BPTIBPTI 33 66 H-H- SSSSSS SS -OH-OH BPTIBPTIBPTI 333 666 H-H- SSSS SS -OH-OH-OH C5C5 C14C14 C30 C38SS C51 C55 BPTIBPTI 33 66 H-H- C5C5 C14C14 SS C30C30C30 C38 C38 C38 C51C51C51 C55C55 -OH-OH BPTI 3 6 H-H- C5 C14 C30 C38 C51 C55 -OH-OH C5C5 C14C14 C30C30 C38 C38 C51C51 C55C55 BPTI 3 6 H- SSSS SSSS C30 C38 C55-OH-OH H- C5SSSSSS C14 SSSSSS C30 C38 C51 C55 C5 SSC14 C22C22 C28C30 C28 C39 C38C39 C51 C55 H-H- C5 SSSSC14 SSSSC22 C28C30C39 C38 C51 C55 -OH-OH H-H- SS SSC22 C28 C39C39 -OH-OH-OH hirudinhirudinhirudin 333 666 C6C6 C14C14C16C16 C22 C28 C39 3SS3SS hirudin 3 6 H-H- C6C6SSC14C14 C16C16SS C22 C28 C39 -OH-OH 4.24.2 3SS3SS H- C22 C28SSSS C39 -OH 4.24.2 3SS 3SS hirudinhirudin 33 66 C6C6 C14C14C16C16C22 C28SSSS C39 4.2 4.2 3SS3SS hirudin 3 6 H- C6 C14 C16 SSSS -OH-OH 4.24.2 3SS hirudin 3 6 SSSS 4.2 3SS hirudin 3 6 C6 C14 C16 SSSSSS 4.2 3SS SS SSSS 4.2 SSSS SS SS SSSS H-H- SS SS -OH-OH-OH enterotoxinenterotoxin 33 6 6 H- C6C6 C7 C7 C11C11 C15C15 -OH enterotoxinenterotoxin 33 6 6 H-H- C6C6 C7 C7 C11C11SS C15C15 -OH-OH enterotoxinenterotoxin 33 6 6 H- C6C6 C7 C7 C11C11SSSSC15C15 -OH enterotoxin 3 6 H- C6 C7 C10C10C10 C11SSC15 C18C18C18-OH H- C6 C7 C10 C11 SSSSC15 C18-OH enterotoxin 3 6 C6 C7 C10C10C11 SSC15 C18C18 C6SS C7 C10 SS C18 SSSSSS C10 SS C18 SS SSSS SS C18 C26C26 C58SSC58 C10SSSS C84C84 C110C110C18 C26C26 C58C58SS C84C84 C110C110 -OH H-H- C26C26 C58SS SSSS C84 C110C110 -OH-OH H- C26 C40 C58C65 C72C72 C84 C110 -OH-OH ribonucleaseribonuclease A A 4 4 8 8 C26 C40C40C40 SSC58C65C65C65SSC72C72 C84C95C95 C110 ribonuclease A 4 8 H-H- C40 SS C65SSC72 C95C95 -OH-OH C26 C40SSSSC58C65SSC72C84 C110 -OH ribonucleaseribonucleaseribonuclease A AA 4 4 4 8 8 8 H- C26 C40SSSSC58 C65 C72C84SSSSC95C95 C110 ribonuclease A 4 8 H- C40 SSSS C65 C72 SSSSC95 -OH-OH C40 SS C65 C72 SSSS ribonuclease A 4 8 C40 SS C65 C72SSSS C95SS SS SSSS SS SSSSSS SS SSSS SSSS SS SS SS SSSSSSSS C76SS C94 SS C76C76C76 C94 C94 C94 lysozyme 4 8 H- SS SSC76 C94 -OH-OH 4SS4SS lysozymelysozymelysozyme 444 8 8 8 H-H-H- SSSS SSC76 C94 -OH-OH 4.34.3 4SS4SS lysozyme 4 8 H- C6C6 C30C30 SSSSC64C64 C76 C94 C115C115C127C127-OH 4.34.3 4SS 4SS lysozymelysozymelysozyme 444 8 8 8 H-H- C6C6C6 C30C30C30 C64C64C64 C76 C94 C115C115C115 C127C127C127-OH-OH 4.3 4.3 4SS4SS lysozyme 4 8 H- SS SSSS -OH 4.34.3 4SS lysozyme 4 8 H- C6C6 C30C30 C64C64 C76SSSS C94 C115C115 C127C127-OH 4.3 4SS lysozyme 4 8 H- C6 C30 C64 SSSS C115 C127-OH 4.3 4SS lysozyme 4 8 H- SSSS SS -OH 4.3 4SS C6 C30 C64 SSSSSS SS C115 C127 4.3 SSSS SS SSSS SSSSSS SSSS SS SSSSSSSS C73SSC73 C91 C91 SSSS SSC73C73 C91 C91 αα-lactalbumin-lactalbumin 44 8 8 H-H- SSSS SSC73 C91 -OH-OH αα-lactalbumin-lactalbumin-lactalbumin 444 8 8 8 H-H- SSSS C73C73 C91 C91 -OH-OH-OH α-lactalbumin 4 8 H- C6C6 C28C28 SSSS C61C61 C73C77C77 C91 C111C111C120C120-OH α-lactalbuminαα-lactalbumin-lactalbumin 444 8 8 8 H-H- C6C6 C28C28 C61C61 C73C77 C91 C111C111 C120C120-OH-OH α-lactalbumin 4 8 H- SS SSSS -OH α-lactalbumin 4 8 H- C6C6 C28C28 C61C61C73C77SSC77SS C91 C111C111 C120C120-OH C6 C28 C61 C111 C120-OH α-lactalbumin 4 8 H- C6 C28 C61 C77SSSS C111 C120-OH C6 C28 C61 C77SSSSSS C111 C120 C6 C28 C61 C77SSSSSS C111 C120 C106C106SSSSC119C119 C106C106 SSC119C119 -OH H-H- C106SSC119 -OH-OH H- C106 C119 C160C160-OH-OH OddOdd Cys Cys ββ-lactoglobulin-lactoglobulin 22 5 5 H- C66C66 C106SSC119 C160C160C160-OH 4.44.4 Odd Cys β-lactoglobulin-lactoglobulin 22 5 5 H- C66C66 SSSSSS SHSH C160-OH 4.4 H- C106SSC119SHSH C160C160-OH OddOdd CysCys ββ-lactoglobulin-lactoglobulin 22 5 5 C66C66 C106 SS C121C121 C160 4.44.4 OddOdd Cys Cysβ -lactoglobulinβ-lactoglobulin 22 5 5 H- C66 SS C121C121SHSH C160-OH-OH 4.4 4.4 Odd Cys β-lactoglobulin 2 5 C66 SS C121C121SH C160 4.4 Odd Cys β-lactoglobulin 2 5 C66(SeSe)(SeSe) SS SHC121 C160 4.4 (SeSe)(SeSe)(SeSe) SS C121SH (SeSe)SSSS C121 C6C6 (SeSe)(SeSe)(SeSe)SSSSSS C11C11 C20C20C121 C6 (SeSe)C11 C20 A-chainA-chain H-H- C6C6 SSSS C11C11 C20-OH-OH A-chain H- C6(SeSe)(SeSe)SS C11 C20-OH-OH insulininsulin 33 A-chain H- C6 C7C7 C11 C20-OH insulininsulininsulin 33 a a A-chain H- C6 SSC7 C11 C20 -OH 4/24/2 a aa A-chain H- C6 C7C7 C11 SSC20SS -OH SS SS insulininsulin 33 4/2 SS C7 SS insulin 3 4/2 A-chain SS SS 4/2 H- SS -OH 4/2 a SS a A-chain SS C7 H- (SeSe) -OH insulin 3 a (SeSe) selenoinsulinselenoinsulin 22 + + 1SeSe 1SeSe (SeSe) SSSS

4/2 (SeSe) (SeSe) selenoinsulin 2 + 1SeSe 4/2 C7SS SS SS selenoinsulin 2 + 1SeSe 4/2 a (SeSe) insulinselenoinsulininsulin 3 2 + 1SeSe3 4/2a B-chain H- C7SS SS -OH B-chain H- (SeSe) -OH SS a B-chain H- (SeSe) -OH selenoinsulinselenoinsulin 22 ++ 1SeSe1SeSe 4/24/2 a B-chain H- (SeSe) SS -OH-OH SS 4/2 (SeSe) C7 C19SS selenoinsulin 2 + 1SeSe B-chainB-chain H-H- C7C7C7SS C19C19C19 -OH-OH selenoinsulin 2 + 1SeSe B-chain H- (SeSe) C7 C19

selenoinsulin 2 + 1SeSe (SeSe) Tow-chainTow-chain selenoinsulin 2 + 1SeSe B-chain H- C7C7 C19C19 -OH 4.54.5 Tow-chain B-chain H- C7 SSSS C19 -OH 4.5 Tow-chain B-chain H- C10C10C7 SSSSSS C15C15C19 C24C24 -OH 4.5 Tow-chainTow-chain C10C7 SS C15C19 C24 4.5 4.5 Tow-chain A-chainA-chain H-H- C10C10C7 SSSS C15C15C19 C24C24-OH-OH 4.5 Tow-chain A-chainA-chain H- C10 SS C15 C24-OH-OH 4.5 Tow-chain H- C10 C11C11 C15 C24-OH 4.5 A-chainA-chain H- C10 SSC11 C15 C24-OH aa A-chain H- C10 C11C11C15 SSC24SS-OH SS a SS C11 SS relaxin 2 3 4/2 aa SS SS-OH relaxin 2 3 4/2 a H- SS A-chain SS

relaxin 2 3 4/2 SS C11 relaxin 2 3 4/2 a a A-chain H- SSSS-OH

C11SS SS a SS relaxinrelaxin 22 3 3 4/2 4/2 a H-H- C11SS SS -OH B-chainB-chain SS -OH relaxin 2 3 4/2a a B-chainB-chain H-H- SS -OH-OH-OH a SS C23SS C11C11SS C23 relaxin 2relaxin 2 3 3 4/2 4/2 B-chainB-chain H-H- C11C11 C23C23 -OH-OH B-chain H- C11 C23 -OH C11C11 C23C23 B-chain H- C11 C23 -OH-OH C11 C23

a Numbers of Cys (or Sec; selenocysteine) residues in the A/B chains, respectively. Molecules 2021, 26, 195 3 of 19

2. Historical Background In 1920–30s, protein denaturation processes were elaborately studied by Wu [5]. Then, about a half century ago, Anfinsen [1,2] proposed a fundamental principle of protein folding, on the basis of the pioneering study using bovine pancreatic ribonuclease A (RNase A) as a model protein; namely, a polypeptide chain eventually folds to one unique structure, called a native state. This principle alternatively brought the researchers a rationale of the so-called Anfinsen’s dogma that the structural information of a protein is encoded in the amino acid sequence. In the meantime, Levinthal [3] theoretically found that, if a polypeptide chain searched for the specific native structure in a trial and error manner, it should require an astronomical time to fold to a protein. Since the protein folding process practically completes within a second for small globular proteins, the big time- scale gap between the theorical prediction and the experimental observation led him to a paradoxical advocation that proteins should fold through their defined pathways. Thus, in an early stage of the protein folding study, the researchers focused on characterizing the folding pathways in order to decode the information of the native-state 3D structure from the amino acid sequence. In the meantime, it became to be widely recognized in around 1990s that neurodegenerative disorders, such as Creutzfeldt–Jakob disease (CJD), Alzheimer’s disease (AD), Parkinson’s disease (PD), amyotrophic lateral sclerosis (ALS), etc., are related to misfolding of proteins in neurons [6,7], suggesting that the native structure of a protein is not a unique state for the amino acid sequence but one of the possible stable states. Moreover, genetic engineering and chemical peptide synthesis technologies have become available for preparation of designed artificial proteins in the last few decades [8]. With these breakthrough events, the protein folding study is now moving to a new realm where the structures of proteins can be actively controlled. Thus, development of the methods to prevent or cure protein misfolding diseases, as well as fabrication of peptide-based medicines, is not our dream. To reach the goal, a proper understanding of the peptide and protein folding pathways depending on the environment must be especially important [9].

3. Physicochemical Aspects of SS-Coupled Protein Folding To characterize the folding pathways of a protein, it is necessary to observe the transiently forming intermediates during the folding reaction [10]. However, in spite of numerous efforts in the past, observation of the folding intermediates is still not easy due to their inherent conformational flexibility and short lifetime. On the other hand, in the oxidative folding of SS-containing proteins, the reaction generally proceeds on a longer time scale (mins to hours). Indeed, the intermediates (i.e., SS intermediates) generated during the oxidative protein folding can be rather easily characterized by trapping them either by acidification of the reaction solution to pH 2 to 3, which converts the reactive cysteinyl thiolates (S−) to essentially unreactive thiols (SH) [11,12], or by chemically cap- ping the S− with quenching reagents, such as iodoacetic acid or iodoacetamide [13]. The SS-intermediates thus inactivated are subsequently separated by HPLC, and their chemical structures, i.e., the SS topologies and conformations, are determined by spectroscopic methods as well as proteolytic digestion. In this way, the oxidative folding pathways of many SS-containing proteins have been proposed. Scheraga and co-workers previously categorized the oxidative folding pathways into four typical types (Figure1)[ 14]. In each type, a reduced and denatured protein (R) folds through a des intermediate, which is generated by SS rearrangement of partially and randomly oxidized species, i.e., the unstructured precursors (see Figure1). The des intermediates represent specific SS intermediates lacking one of the native SS bonds present in the native state. The desU has no specific structure, whereas the desN adopts a native- like structure. These des intermediates can be converted to the native state (N) when the direct oxidation to form the last native SS bond is possible. However, some of the desN intermediates cannot directly be oxidized to N due to the stiff conformation. When a local unfolding is possible in the desN to generate desLU, the buried SH groups would Molecules 2021, 26, x FOR PEER REVIEW 4 of 20

Molecules 2021, 26, 195 4 of 19 oxidation to form the last native SS bond is possible. However, some of the desN intermediates cannot directly be oxidized to N due to the stiff conformation. When a local unfolding is possible in gainthe desN enough to generate ﬂexibility desL to formU, the the buried ﬁnal SS SH linkage groups by would oxidation. gain enough Chang [ 4] delineated flexibility to formthe the relationship final SS linkage between by oxidation. the number Chang of SS bonds[4] delineated and the diversity the relationship of conformations (i.e., between the numberchain of entropy)SS bonds during and the the divers oxidativeity of folding conformations by using (i.e., a folding chain funnelentropy) as the energetic landscape [15]. during the oxidative folding by using a folding funnel as the energetic landscape [15].

Figure 1. Four typical oxidative folding pathways governed by a folding funnel as the energetic landscape. This ﬁgure was madeFigure by merging1. Four typical a concept oxidative regarding folding theenergetic pathways landscape governed model by a folding by Chang, funnel et al. as [the4] to energetic the schematic model demonstratedlandscape. by Scheraga, This figure et al. was [14 ].made by merging a concept regarding the energetic landscape model by Chang, et al. [4] to the schematic model demonstrated by Scheraga, et al. [14]. Figure2 schematically shows the reaction mechanism of SS formation, which is a Figure 2 schematicallyfundamental shows process the of reaction the oxidative mechanism protein folding.of SS formation, To regenerate which N from is a the reduced R, fundamental processSS bonds of the must oxidative be formed protein at the folding. correct To positions regenerate of the Nnative from the state. reduced When R is oxidized, R, SS bonds must however,be formed the at the SS formationcorrect positi usuallyons of occurs the native before state. the When structural R is oxidized, folding. Therefore, in however, the SS formationthis pre-folding usually event, occurs numerous before the intermediates structural folding. having randomlyTherefore, formed in this SS bonds are pre-folding event,generated numerous (Figure intermediates2, Phase I). having Then, the randomly partially formed oxidized SS species bonds gradually are gener- gain the native- ated (Figure 2, Phaselike I). structure Then, the accompanied partially ox byidized regeneration species of gradually the native gain SS pairings the native-like via SS rearrangement (Figure2, Phase II). The latter process is governed by thermodynamics of the peptide structure accompanied by regeneration of the native SS pairings via SS rearrangement chain under the applied conditions. Thus, the formation of the native SS bonds after the (Figure 2, Phase II). The latter process is governed by thermodynamics of the peptide pre-folding event would be coupled with the conformational shift from the random-coil or chain under the applied conditions. Thus, the formation of the native SS bonds after the collapsed state to the native-like fold [16]. pre-folding event wouldIn thebe oxidativecoupled with folding the study,conformational the selection shift of from an oxidant the random-coil (SS-forming agent) is an or collapsed stateimportant to the native-like factor to fold control [16]. the pathways of a protein folding because it greatly affects the reaction rate of SS formation, thereby controlling the amounts of SS intermediates that are transiently accumulated in the reaction solution. In early years, aerial oxidation Molecules 2021, 26, x FOR PEER REVIEW 5 of 20

Molecules 2021, 26, 195 5 of 19

by molecular oxygen (O2) was frequently applied, although the oxidation proceeded very slowly with O2. To perform the oxidative folding at a reasonable reaction velocity, dimeric SS oxidants, such as glutathione (GSSG), a representative oxidant also in vivo, were subsequently employed. As shown in Scheme1a, the dimeric SS oxidant induces formation of a SS bond in a protein via a mixed-SS intermediate. However, generation of such SS intermediates was troublesome in analyzing the folding pathways. In this regard, a use of trans-4,5-dihydroxy-1,2-dithiane (DTTox) is advantageous because the Figure 2. Disulfide formation and reshuffling governed by the thermodynamic stability. mixed-SS intermediate would not be populated in the reaction solution (Scheme1b). This significantly decreases the number of the detectable folding intermediates [17]. However, In the oxidative folding study, the selection of an oxidant (SS-forming agent) is an since oxidation potentials for GSSG and DTTox (E◦0 = −256 and −327 mV, respectively [18]) areimportant close to that factor for cystineto control (CysSSCys) the pathways (E◦0 = −of238 a protein mV [18 ]),folding SS formation because in it a greatly peptide affects the chainreaction should rate be of reversible, SS formation, hencean thereby excess contro amountlling of the the oxidant amounts is required of SS intermediates to complete that are thetransiently reaction. Moreover, accumulated the SS-based in the reaction oxidants solution. are applicable In early only years, under aerial a weakly oxidation basic by molec- Molecules 2021, 26, x FOR PEER REVIEW 5 of 20 conditionular oxygen (pH 7.5~9.0).(O2) was As frequently a consequence, applied, clear although observation the ofoxidation conformational proceeded folding very slowly eventswith coupledO2. To withperform SS rearrangement the oxidative is stillfolding not easyat a evenreasonable with these reaction SS-based velocity, oxidants. dimeric SS oxidants, such as glutathione (GSSG), a representative oxidant also in vivo, were subsequently employed. As shown in Scheme 1a, the dimeric SS oxidant induces formation of a SS bond in a protein via a mixed-SS intermediate. However, generation of such SS intermediates was troublesome in analyzing the folding pathways. In this regard, a use of trans-4,5-dihydroxy-1,2-dithiane (DTTox) is advantageous because the mixed-SS intermediate would not be populated in the reaction solution (Scheme 1b). This significantly decreases the number of the detectable folding intermediates [17]. However, since oxidation potentials for GSSG and DTTox (E°′ = −256 and −327 mV, respectively [18]) are close to that for cystine (CysSSCys) (E°′ = −238 mV [18]), SS formation in a peptide chain should be reversible, hence an excess amount of the oxidant is required to complete the reaction. Moreover, the SS-based oxidants are applicable only under a weakly basic condition (pH 7.5~9.0). As a consequence, clear observation of conformational folding events coupled FigureFigurewith 2. SS 2.Disulfide Disulfiderearrangement formation formation and is andstill reshuffling reshuffling not easy governed evengoverned bywith the by these thermodynamic the thermodynamic SS-based stability. oxidants. stability.

In the oxidative folding study, the selection of an oxidant (SS-forming agent) is an important factor to control the pathways of a protein folding because it greatly affects the reaction rate of SS formation, thereby controlling the amounts of SS intermediates that are transiently accumulated in the reaction solution. In early years, aerial oxidation by molecular oxygen (O2) was frequently applied, although the oxidation proceeded very slowly with O2. To perform the oxidative folding at a reasonable reaction velocity, dimeric SS oxidants, such as glutathione (GSSG), a representative oxidant also in vivo, were subsequently employed. As shown in Scheme 1a, the dimeric SS oxidant induces formation of a SS bond in a protein via a mixed-SS intermediate. However, generation of such SS intermediates was troublesome in analyzing the folding pathways. In this regard, a use of trans-4,5-dihydroxy-1,2-dithiane (DTTox) is advantageous because the mixed-SS intermediate would not be populated in the reaction solution (Scheme 1b). This significantly decreases the number of the detectable folding intermediates [17]. However, since oxidation potentials for GSSG and DTTox (E°′ = −256 and −327 mV, respectively [18]) are close to that for cystine (CysSSCys) (E°′ = −238 mV [18]), SS formation in a peptide chain should be reversible, hence an excess amount of the oxidant is required to complete the reaction. Moreover, the SS-based oxidants are applicable only under a weakly basic condition (pH 7.5~9.0). As a consequence, clear observation of conformational folding events coupled with SS rearrangement is still not easy even with these SS-based oxidants.

SchemeScheme 1. Mechanisms 1. Mechanisms for SS for formation SS formation using varioususing various oxidants. oxidants. (a) A dimeric (a) A disulﬁde dimeric (SS) disulfide reagent. (SS) rea- (b)gent. A cyclic (b) SSA cyclic reagent. SS ( creagent.) A selenoxide (c) A reagentselenoxide [19]. reagent [19].

Scheme 1. Mechanisms for SS formation using various oxidants. (a) A dimeric disulfide (SS) reagent. (b) A cyclic SS reagent. (c) A selenoxide reagent [19]. Molecules 2021, 26, x FOR PEER REVIEW 6 of 20

Molecules 2021, 26, 195 6 of 19 To separate the SS formation and SS rearrangement events during oxidative protein folding, trans-3,4-dihydroxytetrahydroselenophene-1-oxide (DHSox) was developed as a strongTo SS-forming separate the reagent SS formation [19,20], andwhich SS enables rearrangement rapid but events selective during oxidation oxidative of the protein SH groupsfolding, of trans-3,4-dihydroxytetrahydroselenophene-1-oxide a reduced protein molecule (Scheme 1c). The reaction (DHS proceedsox) was developedstoichiometri- as a callystrong in SS-forminga wide range reagent of pH [19 (i.e.,,20], pH which 3.0~10.0) enables through rapid buta similar selective mechanism oxidation to of that the SHof DTTgroupsox. In of the a reduced first step, protein the SH molecule group (Schemeof a pept1c).ide The chain reaction attacks proceeds at the Se stoichiometrically atom of DHSox. Then,in a wide another range SH of group pH (i.e., of the pH peptide 3.0~10.0) chain through intramolecularly a similar mechanism attacks at to the that S ofatom DTT ofox the. In generatedthe first step, thioselenurane the SH group intermediate of a peptide chainto produce attacks a at SS the bond, Se atom releasing of DHS selenideox. Then, DHS anotherred. SinceSH group the oxidation of the peptide potential chain of DHS intramolecularlyox (E°′ = 375 mV attacks [21]) is at remarkably the S atom ofhigher the generatedthan that ofthioselenurane cystine, the rate-determining intermediate to step produce becomes a SS the bond, first releasing step and selenidethe reaction DHS proceedsred. Since irre- the ox ◦0 versiblyoxidation and potential stoichiometrically. of DHS (E According= 375 mV to [21 these]) is remarkably characteristic higher features than thatof DHS of cystine,ox, the reactionthe rate-determining rate of the SS formation step becomes was theproporti first steponal andto the the number reaction of free proceeds SH groups irreversibly exist- ox ingand in stoichiometrically. the substrate peptide According chain [19,21,22] to these. characteristicFurthermore, featuresthe SS formation of DHS process, the reaction com- pletedrate of before the SS the formation SS reshuffling was proportional in the oxidized to the numberpeptide ofchain. free SHThus, groups Phases existing I and inII theof Figuresubstrate 2 could peptide be chainseparated, [19,21 ,making22]. Furthermore, the analysis the of SS the formation oxidative process folding completed pathways before of proteinsthe SS reshuffling much easier. in the In oxidizedaddition peptideto DHSox chain., inorganic Thus, oxidants, Phases I and such II as of a Figure platinum2 could com- be plexseparated, ([Pt(en) making2Cl2]2+), thewere analysis also employed of the oxidative [23–25]. folding Various pathways diselenide of (SeSe)- proteins and much SS-based easier. ox 2+ reagentsIn addition (or catalysts), to DHS ,which inorganic can accelerate oxidants, suchthe oxidative as a platinum folding complex velocity, ([Pt(en) were tested2Cl2] as), alternativewere also employed oxidants (see [23– [26]25]. for Various a review diselenide and [27–34] (SeSe)- for and recent SS-based papers). reagents In the (orlast catalysts), decade, thewhich enzymes can accelerate that assist the protein oxidative folding folding processes velocity, in the were endoplasmic tested as reticulum alternative (ER), oxidants such as(see protein [26] for disulfide a review isomerase and [27– 34(PDI),] for recentwere fu papers).rther applied In the lastto enhance decade, both the enzymes the velocity that andassist yield protein of oxidative folding processes protein folding in the endoplasmic practically [35]. reticulum (ER), such as protein disulfide isomerase (PDI), were further applied to enhance both the velocity and yield of oxidative 4.protein Oxidative folding Folding practically of Peptides [35]. and Proteins 4.1.4. Oxidative Two-Disulfide Folding Peptides of Peptides and Proteins 4.1. Two-DisulfideWhen a reduced Peptides peptide chain (R) having four cysteinyl SH groups is oxidized by a suitableWhen oxidant, a reduced a number peptide of heterogeneou chain (R) havings SS-intermediates four cysteinyl SHwould groups be generated, is oxidized in- by cludinga suitable theoretically oxidant, aup number to six one-disulfide of heterogeneous (1SS) SS-intermediatesand three two-disulfide would (2SS) be generated, intermediates.including Indeed, theoretically sequential up formation to six one-disulfide of the ensembles (1SS) andof 1SS three and two-disulfide 2SS intermediates (2SS) were inter- observedmediates. when Indeed, reduced sequential A-chains formation of bovine of theinsulin ensembles (Ins-A) ofand 1SS human and 2SS relaxin intermediates 2 (Rlx-A) werewere oxidized observed with when stoichiometric reduced A-chains amounts of of bovine DHSox insulin as shown (Ins-A) in eq anduation human (1) [21]. relaxin 2 (Rlx-A) were oxidized with stoichiometric amounts of DHSox as shown in Equation (1) [21].

(1)

TheThe kinetic kinetic analyses analyses clearly clearly showed showed th thatat the the second second order order rate rate constants, constants, k1 andk1 and k2, whichk2, which correspond correspond to transformations to transformations of R of→ R1SS→ and1SS 1SS and → 1SS 2SS,→ respectively,2SS, respectively, were werepro- portionalproportional to the to number the number of free of SH free groups SH groups existing existing along along the reactant the reactant peptide peptide chains chains (i.e., k(i.e.,1: k2 k=1 2:1).: k2 = The 2:1). observed The observed stochastic stochastic SS format SS formationion suggested suggested that thatthe SS the intermediates SS intermediates of Ins-Aof Ins-A and and Rlx-A Rlx-A have have fully fully flexible ﬂexible conformations. conformations. OnOn the the other other hand, when thethe reducedreduced species species of of endothelin-1 endothelin-1 (ET-1), (ET-1), a vasoconstrictivea vasoconstric- tivepeptide peptide consisting consisting of 21 of amino 21 amino acid residues, acid residues, was oxidized was oxidized with O 2with, only O two2, only 2SS productstwo 2SS productsout of three out wereof three obtained were obtained in a ratio in of a 1: ratio 3. The of 1: minor 3. The isomer minor had isomer non-native had non-native SS pairings SS pairings(i.e., Cys1–Cys11 (i.e., Cys1– and Cys11 Cys3–Cys15), and Cys3– while Cys15), the while other the major other one major had twoone nativehad two SS native bonds SS(i.e., bonds Cys1–Cys15 (i.e., Cys1– and Cys15 Cys3–Cys11) and Cys3– (Figure Cys11)3)[ (Figure36]. Moroder 3) [36]. Moroder et al. replaced et al. replaced two seleno- two selenocysteinecysteine (Sec) (Sec) residues residues for the for two the innertwo inner Cys residuesCys residues in order in order to selectively to selectively regenerate regen- eratenative native ET-1. ET-1. This This SS-to-SeSe SS-to-SeSe strategy strategy relies relies on the on lower the lower reduction reduction potential potential of a SeSe of a bondSeSe (E◦0 = −380 ~ −410 mV [37]) than that of a SS bond (E◦0 = –230 ~ –260 mV [37]), in- bondSeSe (E°′SeSe = −380 ~ −410 mV [37]) than that of a SS bond (ESS°′SS = –230 ~ –260 mV [37]), indicatingdicating that that a a SeSe SeSebond bond isis thermodynamicallythermodynamically more stable than a SS SS bond. bond. Moreover, Moreover, the pKa value of the selenol (SeH) group in Sec is about 3 units lower than that of the SH the pKa value of the selenol (SeH) group in Sec is about 3 units lower than that of the SH group. Thus, the SeH group is more reactive than thiols, and, therefore, the formation of a group. Thus, the SeH group is more reactive than thiols, and, therefore, the formation of SeSe bond does not compete with the formation of a SS bond, facilitating the peptide chain a SeSe bond does not compete with the formation of a SS bond, facilitating the peptide to fold to the correct structure. In fact, the ET-1 analogue, [Sec3, Sec11, Nle7]-ET-1, selec- chain to fold to the correct structure. In fact, the ET-1 analogue, [Sec3, Sec11, Nle7]-ET-1, tively produced the SeSe bond, yielding the native folded state exclusively (Figure3) . The obtained SeSe analogue of ET-1 showed almost the same biological activity as native ET-1. Molecules 2021, 26, x FOR PEER REVIEW 7 of 20

selectively produced the SeSe bond, yielding the native folded state exclusively (Figure

Molecules 2021, 26, 195 3). The obtained SeSe analogue of ET-1 showed almost the same biological7 of activity 19 as native ET-1.

Natural oxidative folding pathway

3 1 Oxidation NH2 X=S 3 1 1 3 HX SH S S S S XH SH S S S S + 11 15 11 15

CO2H 11 15 25% 75% Native endothelin-1 X=Se

3 1 3 1

NH2 NH2 Se SH Se S Se SH Se S

CO2H CO2H 11 15 11 15 100% Modified oxidative folding pathway

FigureFigure 3. Oxidative 3. Oxidative folding folding pathways pathways of [Sec of 3[Sec, Sec113, ,Sec Nle117]-endothelin-1, Nle7]-endothelin-1 (ET-1) via predominant(ET-1) via pred diselenideominant (SeSe)-bond diselenide formation. (SeSe)-bond formation. The SS-to-SeSe strategy is useful not only for obtaining correctly folded proteins in high yields but also for steering the folding pathways. Moroder et al. successfully applied theThe strategy SS-to-SeSe to the oxidative strategy folding is useful of apamin, not only a small for bee obtaining venom peptide correctly comprised folded of proteins in high18 yields amino acidbut residuesalso for having steering two the SS bondsfoldin (Cys1–Cys15g pathways. and Moroder Cys3–Cys11) et al. in successfully the native applied the state.strategy They to prepared the oxidative the all possible folding isomers, of apamin, i.e., [Sec1–Sec11, a small Cys3–Cys15],bee venom [Sec3–Sec11,peptide comprised of Cys1–Cys15], and [Sec1–Sec3, Cys11–Cys15], and analyzed the relationship between the 18 amino acid residues having two SS bonds (Cys1–Cys15 and Cys3– Cys11) in the native peptide conformations and the topologies of the SS/SeSe bonds by means of CD and state.NMR They spectroscopy prepared [38 the]. Similar all possible strategies isomers, have been i.e., widely [Sec1–Sec11, applied to Cys3–Cys15], various peptides [Sec3–Sec11, Cys1–Cys15],and proteins and having [Sec1–Sec3, multiple SS Cys11–Cys15], bond in the native and state, analyzed such as conotoxine, the relationship Ecballium between the peptideelaterium conformations trypsin inhibitor and II (EETI-II),the topologies and bovine of the pancreatic SS/SeSe trypsin bonds inhibitor by means (BPTI) of (vide CD and NMR spectroscopyinfra) [39–45 ].[38]. Similar strategies have been widely applied to various peptides and The success of the SS-to-SeSe strategy demonstrated that the thermodynamic stability proteinsof a SeSe having bond can multiple change the SS folding bond pathwaysin the native of peptides state, and such proteins. as conotoxine, Thus, intrinsically Ecballium elate- riumnon-productive trypsin inhibitor folding intermediatesII (EETI-II), can and be purposelybovine pancreatic produced and tr isolatedypsin inhibitor by replacing (BPTI) (vide infra)Sec [39–45]. for the Cys residues.

4.2.The Three-Disulﬁde success of Proteinsthe SS-to-SeSe strategy demonstrated that the thermodynamic stability of a SeSe bond can change the folding pathways of peptides and proteins. Thus, intrinsi- BPTI is a 58-residue globular protein, having three SS linkages (i.e., Cys5–Cys55, callyCys14–Cys38, non-productive and Cys30–Cys51) folding intermediates in the native folded can state.be purposely This small protein produced was one and of isolated by replacingthe most Sec extensively for the studied Cys residues. model proteins in the protein folding ﬁeld before 2000s [46–51]. Following the pioneering works of Creighton [46,47], Weissman and Kim [48–50] unam- 4.2.biguously Three-Disulfide determined Proteins the oxidative folding pathways as shown with black arrows in Figure4. All the important SS intermediates were characterized in terms of their SS topolo- giesBPTI as well is asa kinetic58-residue and thermodynamic globular protein, stabilities. having In this conventionalthree SS linkages folding pathways, (i.e., Cys5–Cys55, Cys14–Cys38,native BPTI (N) and is slowlyCys30–Cys51) regenerated in from the two native 2SS intermediates, folded state. indicated This small with symbolsprotein was one of 0 the N*most and extensively N , which have studied two native model SS proteins bonds between in the the protein Cys residues folding denoted field before in the 2000s [46– boxes. These intermediates are highly structured, thereby being kinetically trapped on the 51]. Following the pioneering works of Creighton [46,47], Weissman and Kim[48–50] unfolding pathways. ambiguously determined the oxidative folding pathways as shown with black arrows in Figure 4. All the important SS intermediates were characterized in terms of their SS topologies as well as kinetic and thermodynamic stabilities. In this conventional folding pathways, native BPTI (N) is slowly regenerated from two 2SS intermediates, indicated with symbols N* and N′, which have two native SS bonds between the Cys residues denoted in the boxes. These intermediates are highly structured, thereby being kinetically trapped on the folding pathways. Molecules 2021, 26, x FOR PEER REVIEW 8 of 20

To detour the sluggish folding pathways of BPTI, Hilvert and Metanis synthesized an artificial (C5U,C14U)-BPTI variant, in which Cys5 and Cys14 are replaced by Sec residues, by applying the solid-phase peptide synthesis methodology [52,53]. The oxidative Molecules 2021, 26, x FOR PEER REVIEW 8 of 20 folding of the obtained variant was indeed accelerated due to appearance of an alternative pathway (Figure 4, red arrows), which goes through the new intermediates having a SeSe bond between Sec5 and Sec14 residues. Thus, the variant can avoid being trapped as N* SH and NTo′ anddetour go to the N SHsluggish directly. folding It is of pathwaysinterest that of the BPTI, (C5U,C14U)-BPTI Hilvert and Metanis variant synthesizeddoes not anhave artificial a SeSe (C5U,C14U)-BPTIbond but has two SeS variant, bonds inin whichthe native Cys5 folded and stateCys14 even are thoughreplaced a SeS by bondSec resi- Molecules 2021, 26, 195 8 of 19 dues,is thermodynamically by applying the lesssolid-phase stable than peptide a SeSe synthesisbond. This methodology may reflect the [52,53]. larger Theconforma- oxidative foldingtional stability of the obtained of N than variant that of wasthe (Sec5–Sec14,indeed acce leratedCys30–Cys51) due to appearanceintermediate. of an alternative pathway (Figure 4, red arrows), which goes through the new intermediates having a SeSe bond between Sec5 and Sec14 residues. Thus, the variant can avoid being trapped as N* SH and N′ and go to NSH directly. It is of interest that the (C5U,C14U)-BPTI variant does not have a SeSe bond but has two SeS bonds in the native folded state even though a SeS bond is thermodynamically less stable than a SeSe bond. This may reflect the larger conformational stability of N than that of the (Sec5–Sec14, Cys30–Cys51) intermediate.

FigureFigure 4. 4.Oxidative Oxidative folding folding pathways pathways of wild-type of wild-type bovine bo pancreaticvine pancreatic trypsin inhibitortrypsin inhibitor (BPTI) (black (BPTI) arrows)(black arrows) and BPTI and variant BPTI (C5U/C14U) variant (C5U/C14U) (red arrows). (red arrows).

ToAnother detour the interesting sluggish folding attempt pathways to steer of the BPTI, folding Hilvert pathways and Metanis of synthesizedBPTI was recently an re- artiﬁcialported (C5U,C14U)-BPTIby Metanis’ group variant, [54]. inThey which synthesi Cys5 andzed Cys14 a novel are replacedBPTI analogue by Sec residues, (i.e., MT-BPTI), byin applyingwhich solvent-exposed the solid-phase peptide Cys14 synthesis and Cys38 methodology residues are [52 ,53connected]. The oxidative via a methylene folding thio- of the obtained variant was indeed accelerated due to appearance of an alternative pathway acetal bond (S–CH2–S) that cannot be cleaved reductively, in order to arrest the major (Figure4, red arrows), whichSH goes through the new intermediates having a SeSe bond 0 betweenFigurefolding 4. Sec5 pathwaysOxidative and Sec14 foldingvia residues.NSH pathways. Contrary Thus, of the to wild-type variantthe expectation can bo avoidvine pancreatic beingthat the trapped generation tryp assin N* inhibitor and of folded N (BPTI) MT- (black arrows)SH and BPTI variant (C5U/C14U)SH (red arrows). andBPTI go should to NSH directly.be prohibited It is of because interest that NSH the, a (C5U,C14U)-BPTIdirect precursor to variant N, cannot does not be haveformed, MT- aBPTI SeSe folded bond but via has a mixed-SS two SeS bonds intermediate in the native with folded GSH state (i.e., even N*-SG), though which a SeS would bond isbe formed thermodynamicallyby theAnother reaction interesting of less N* stablewith attempt thanGSSG a SeSe(Figure to bond.steer 5). Thisthe Thus, mayfolding the reﬂect introduction pathways the larger conformationalof of BPTIthe uncleavable was recently SS re- stability of N than that of the (Sec5–Sec14, Cys30–Cys51) intermediate. portedbond unveiled by Metanis’ a new group pathway [54]. thatThey had synthesi been hiddenzed a novel in the BPTI previous analogue oxidative (i.e., MT-BPTI),folding in whichAnother solvent-exposed interesting attempt Cys14 to steer and the Cys38 folding residues pathways are of connected BPTI was recentlyvia a methylene re- thio- portedstudies by of Metanis’ wild-type group BPTI. [54]. They synthesized a novel BPTI analogue (i.e., MT-BPTI), in whichacetal solvent-exposed bond (S–CH2 Cys14–S) that and cannot Cys38 residues be cleaved are connected reductively, via a methylene in order thioacetalto arrest the major SH bondfolding (S–CH pathways2–S) that via cannot NSH be. Contrary cleaved reductively, to the expectation in order to that arrest the the generation major folding of folded MT- SH SH pathways via NSH. Contrary to the expectation that the generation of folded MT-BPTI BPTI should be prohibited becauseSH NSH, a direct precursor to N, cannot be formed, MT- should be prohibited because NSH, a direct precursor to N, cannot be formed, MT-BPTI foldedBPTI folded via a mixed-SS via55 a mixed-SS intermediate intermediate with GSH (i.e., with N*-SG), GSH which (i.e., wouldN*-SG), be which formed would by the be formed reactionby the reaction of N* with of GSSG N* with (Figure GSSG5). Thus, (Figure the introduction5). Thus, the of introduction the uncleavable of SS the bond uncleavable SS unveiledbond unveiled30 a new pathway a new that pathway had been that hidden had× in been the previous hidden oxidative in the previous folding studies oxidative of folding wild-type14 BPTI. studies of38 wild-type51 BPTI. 5

MT BPTI × Cys17 S S Cys38 55 Methylene30 thioacetal bridge × 14 38 51 Figure5 5. Oxidative folding pathway of MT-BPTI through direct conversion of N* intermediates.

MT BPTI × Cys17 S S Cys38

Methylene thioacetal bridge

Figure 5. 5.Oxidative Oxidative folding folding pathway pathway of MT-BPTI of MT-BPTI through through direct conversion direct conversion of N* intermediates. of N* intermediates. Molecules 2021, 26, x FOR PEER REVIEW 9 of 20

Another well-studied small 3SS protein is hirudin, a thrombin inhibitor, that has the Molecules 2021, 26, 195 SS linkages between Cys6–Cys14, Cys16–Cys28, and Cys22–Cys39. Although9 ofthe 19 molecular size is similar to BPTI, the oxidative folding pathways of hirudin are in significant contrast Anotherto those well-studied of BPTI. While small 3SSonly protein five ou is hirudin,t of 75 possible a thrombin SS inhibitor, intermediates that has thewere SS located on thelinkages oxidative between folding Cys6–Cys14, pathways Cys16–Cys28, of BPTI (Figure and Cys22–Cys39. 4, black arrows), Although more the molecular than 30 hetero- geneoussize is SS similar intermediates to BPTI, the oxidativecould be foldingdetected pathways for hirudin. of hirudin Chang are in [4] significant reported contrast that ensem- blesto of those the scrambled of BPTI. While SS onlyintermediates five out of 75containi possibleng SS one, intermediates two, and werethree located SS bonds on the (1SS, 2SS, oxidative folding pathways of BPTI (Figure4, black arrows), more than 30 heterogeneous and 3SS, respectively) were generated when reduced hirudin (R) was oxidized by O2 dis- SS intermediates could be detected for hirudin. Chang [4] reported that ensembles of the solvedscrambled in a buffer SS intermediates solution in containing the presence one, two,of β-mercaptoethanol and three SS bonds (1SS,as a 2SS,thiol-based and 3SS, folding catalyst.respectively) This pre-folding were generated process when would reduced correspond hirudin (R) wasto a oxidized hydrophobic by O2 dissolved collapse in of a random-coila buffer state solution of the in peptide the presence chain of βto-mercaptoethanol lose the chain entropy. as a thiol-based The generated folding catalyst. 3SS interme- diatesThis were pre-folding subsequently process converted would correspond to native to ahirudin hydrophobic (N) through collapse ofthe a intermolecular random-coil SS– SH exchangestate of the with peptide β-mercaptoethanol chain to lose the chain in the entropy. solution The (Figure generated 6, blue 3SS intermediates arrow) [55–57]). In were subsequently converted to native hirudin (N) through the intermolecular SS–SH the meantime, Scheraga et al. reported that when DTTox was used as an oxidant, the spe- exchange with β-mercaptoethanol in the solution (Figure6, blue arrow) [ 55–57]). In the cificmeantime, 2SS intermediates Scheraga et that al. reported had two that native when DTTSS bondsox was usedand aswere an oxidant, in equilibrium the specific with the other2SS heterogeneous intermediates that 2SS had intermediates, two native SS were bonds ox andidized were to in 3SS*, equilibrium which with had the three other native SS bondsheterogeneous but did not 2SS yet intermediates, get the native were structure. oxidized toThe 3SS*, generated which had 3SS* three finally native folded SS bonds to N rap- idlybut with did a not gain yet of get the the conformation native structure. stability The generated (Figure 3SS* 6, finally red arrows, folded to [58]). N rapidly Thus, with the oxidative afolding gain of the pathways conformation of hirudin stability (Figurewould6 ,be red switchable arrows, [ 58]). by Thus, selection the oxidative of the folding oxidant em- pathways of hirudin would be switchable by selection of the oxidant employed. ployed.

Figure 6.Figure Revised 6. Revised oxidative oxidative folding folding pathways pathways of ofhirudin, hirudin, combinin combiningg the the models models proposed proposed by Chang by Chang (blue arrow)(blue arrow) [55–57], [55–57], ScheragaScheraga (red arrows) (red arrows) [58], [and58], andArai Arai & Iwaoka & Iwaoka [19,21]. [19,21]. Later on, the early folding events of a recombinant hirudin (CX-397), which has a hybridLater sequenceon, the early of hirudin folding variants-1 events and of -3a [recombinant59], were reinvestigated hirudin (CX-397), using DHS whichox as has a hybridan oxidant sequence [19, 21of]. hirudin The second-order variants-1 rate and constants -3 [59], for were R → reinvestigated1SS, 1SS → 2SS, using and 2SS DHS→ ox as an oxidant3SS (i.e.,[19,21].k1, k 2The, and second-orderk3, respectively) rate were constants roughly infor a ratioR → of1SS, 3:2:1, 1SS being → 2SS, proportional and 2SS → 3SS to the number of the SH groups present in the reactant (see Section3), under acidic to (i.e., k1, k2, and k3, respectively) were roughly in a ratio of 3:2:1, being proportional to the neutral pH conditions (Table2). However, under weakly basic conditions, the ratio became number of the SH groups present in the reactant (see Section 3), under acidic to neutral signiﬁcantly deviated due to the conformational shift from the stochastically formed SS pH intermediatesconditions (Table (1SS◦, 2SS2). ◦However,, and 3SS◦) tounder the thermodynamically weakly basic conditions, stabilized SS-intermediates the ratio became significantly(1SS, 2SS, deviated and 3SS). due In to fact, the in conformational the presence of shift guanidinium from the chloride stochastically (Gdn-HCl) formed as a SS in- termediatesdenaturant, (1SS°, the ratio2SS°, of and 3:2:1 3SS°) was almostto the recoveredthermodynamically although the stabilized overall reaction SS-intermediates was (1SS,decelerated 2SS, and 3SS). due probably In fact, toin the the interactions presence of between guanidinium a guanidinium chloride ion and(Gdn-HCl) the peptide as a dena- ◦ turant,molecule. the ratio Formation of 3:2:1 of awas more almost compact recover structureed inalthough 3SS than inthe 3SS overallwas supported reaction bywas the deceler- ﬂuorescence measurement. Thus, the presence of kinetic and thermodynamic phases in the ated due probably to the interactions between a guanidinium ion and the peptide mole- early SS formation event was elucidated by using DHSox as a selective and strong oxidant. cule. Formation of a more compact structure in 3SS than in 3SS° was supported by the fluorescence measurement. Thus, the presence of kinetic and thermodynamic phases in the early SS formation event was elucidated by using DHSox as a selective and strong oxidant. Molecules 2021, 26, 195 10 of 19

Table 2. The second-order rate constants for oxidation (SS formation) of hirudin CX-397 [19,21].

k for R → 1SS k for 1SS → 2SS k for 2SS → 3SS pH and Additive 1 2 3 (mM−1 s−1) (mM−1 s−1) (mM−1 s−1) 4.0 4.3 ± 0.2 2.6 ± 0.1 1.2 ± 0.1 7.0 9.8 ± 0.5 6.6 ± 0.3 2.8 ± 0.3 8.0 18.6 ± 1.0 10.0 ± 0.6 3.6 ± 0.6 8.0, +2 M Gdn-HCl 9.8 ± 0.6 5.0 ± 0.3 2.7 ± 0.2

Very recently, Shimamoto and Hidaka [60] reported on the folding of a topological isomer of enterotoxin, which is a 3SS short polypeptide produced by enterotoxigenic E. coli. The topological isomer, possessing the three native SS bonds but with a different folded structure, was obtained by regioselective sequential SS-bond formation, while under an aerial oxidation condition only native enterotoxin was generated. This is an interesting example demonstrating that the oxidative folding pathways of a protein are ﬂexible depending on the conditions. The misfolded enterotoxin could be rescued to the native structure in the presence of an external thiol like the case of the 3SS intermediate ensemble of hirudin.

4.3. Four-Disulfide Proteins Bovine pancreatic ribonuclease A (RNase A) has been a symbolic target of biological studies for a long time. This well-characterized globular protein contains four SS linkages, two of which are formed between Cys40–Cys95 and Cys65–Cys72 stabilizing the flexible loop domains and the other two are formed between Cys26–Cys84 and Cys58–Cys110 connecting the α-helix and β-sheet regions in the native state. The oxidative folding pathways of RNase A were well characterized by Scheraga and co-workers using SS-based oxidants, such as DTTox and GSSG [17,61,62]. In the early folding event, reduced RNase A (R) was sequentially oxidized to SS-intermediate ensembles (i.e., 1SS, 2SS, 3SS, and 4SS) having no specific structure. The generated 3SS intermediates were subsequently rearranged to thermodynamically stable des species, which have three native SS bonds but lack one native SS bond denoted in brackets (Figure7). This rate-determining SS- rearrangement process occurred with conformational folding of the native-like structure. At 25 ◦C, only des[40-95] and des[65-72] were observed as the key folding intermediates, which could be oxidized to the native state (N) [61]. On the other hand, des[26-84] and des[58-110], which were observed only at the temperature below 15 ◦C, were dead-end Molecules 2021, 26, x FOR PEER REVIEWspecies and could not be directly oxidized to N [62]. However, characterization11 of of20 these

des intermediates was not easy in the presence of excess amounts of the oxidants.

FigureFigure 7. Oxidative 7. Oxidative folding folding pathways pathways of of RNase RNase A, A, combiningcombining the the mo modelsdels proposed proposed by byScheraga, Scheraga, et al. et[61,62] al. [61 and,62 Iwaoka,] and Iwaoka, et al. [et63 al.,64 [63,64].]. To monitor the oxidative folding pathways of RNase A more clearly, we applied DHSox as an alternative oxidant [22]. Interestingly, while a use of DHSox did not change the oxidative folding pathways, it revealed the presence of kinetic and thermodynamic phases, which corresponds to the stochastic SS formation with a loss of the chain entropy and the rapid SS reshuffling with a gain of the slight stabilization due to the hydrophobic collapse, respectively, in the pre-folding event [63]. In addition, since DHSox enabled stoichiometric SS formation (Scheme 1c) and no oxidant was left in the reaction solution, the rate-determining steps became the final oxidation processes of des[40-95] and des[65-72] to N, resulting in accumulation of these des species. Hence, the conformational folding processes of RNase A, i.e., 3SS → des intermediates, could be cleanly monitored. After the oxidation of R with DHSox, a pseudo-equilibrium was indeed attained between 3SS and the des intermediates. Therefore, the thermodynamic stability (∆G) of the des intermediates could be estimated from the equilibrium constants [64]. Interestingly, while des[40- 95] was more stable than des[65-72] at 25 °C (∆∆G = −0.7 kcal/mol), the relative stability was inverted at 35 °C (∆∆G = 0.4 kcal/mol). Thus, at the temperature close to a physiological condition, des[65-72] would play a more important role than des[40-95] as a key folding intermediate leading to native RNase A. A temperature effect on the oxidative folding pathways of a protein was more obvious in the case of hen egg-white lysozyme (HEL) (Figure 8a). Dobson et al. [65] reported that the heterogeneous intermediate ensembles, 1SS and 2SS, with no specific structure were generated from reduced HEL (R) in the early folding event at 20 °C (Figure 8b). Then, the 2SS was slowly oxidized to three des intermediates, des[6-127], des[64-80], and des[76- 94], having three native SS linkages. Finally, by oxidation of the two Cys residues left in the des intermediates, native HEL (N) was regenerated. While the oxidation steps of des[64–80] and des[6-127] to N shared the major folding pathways of HEL, the oxidation of des[76-94] was hindered and occurred only with a long reaction time or in the presence of PDI through the other two des intermediates [65,66]. This classical scenario has been revised recently by applying DHSox as an oxidant [67]. When R was reacted with stoichiometric amounts of DHSox, not only 1SS and 2SS but also 3SS and 4SS were generated in a similar manner that was observed for hirudin and RNase A (see above). Then, the 3SS was gradually converted to the three des intermediates via slow SS rearrangement (Figure 8b). More importantly, des[64-80] and des[6-127], which were located on the major folding pathways at 25 °C and below, were not formed Molecules 2021, 26, 195 11 of 19

To monitor the oxidative folding pathways of RNase A more clearly, we applied DHSox as an alternative oxidant [22]. Interestingly, while a use of DHSox did not change the oxidative folding pathways, it revealed the presence of kinetic and thermodynamic phases, which corresponds to the stochastic SS formation with a loss of the chain entropy and the rapid SS reshuffling with a gain of the slight stabilization due to the hydrophobic collapse, respectively, in the pre-folding event [63]. In addition, since DHSox enabled stoichiometric SS formation (Scheme1c) and no oxidant was left in the reaction solution, the rate-determining steps became the final oxidation processes of des[40-95] and des[65-72] to N, resulting in accumulation of these des species. Hence, the conformational folding processes of RNase A, i.e., 3SS → des intermediates, could be cleanly monitored. After the oxidation of R with DHSox, a pseudo-equilibrium was indeed attained between 3SS and the des intermediates. Therefore, the thermodynamic stability (∆G) of the des intermediates could be estimated from the equilibrium constants [64]. Interestingly, while des[40-95] was more stable than des[65-72] at 25 ◦C(∆∆G = −0.7 kcal/mol), the relative stability was inverted at 35 ◦C(∆∆G = 0.4 kcal/mol). Thus, at the temperature close to a physiological condition, des[65-72] would play a more important role than des[40-95] as a key folding intermediate leading to native RNase A. A temperature effect on the oxidative folding pathways of a protein was more obvious in the case of hen egg-white lysozyme (HEL) (Figure8a). Dobson et al. [ 65] reported that the heterogeneous intermediate ensembles, 1SS and 2SS, with no specific structure were generated from reduced HEL (R) in the early folding event at 20 ◦C (Figure8b). Then, the 2SS was slowly oxidized to three des intermediates, des[6-127], des[64-80], and des[76-94], having three native SS linkages. Finally, by oxidation of the two Cys residues left in the des intermediates, native HEL (N) was regenerated. While the oxidation steps of des[64-80] and des[6-127] to N shared the major folding pathways of HEL, the oxidation of des[76-94] was hindered and occurred only with a long reaction time or in the presence of PDI through the other two des intermediates [65,66].

Figure 8. Oxidative folding pathways of hen egg-white lysozyme (HEL) and αLA. (a) 3D structures of HEL (left) and αLA (right). (b) Oxidative folding pathways of HEL. The red and blue arrows represent pathways proposed by Dobson et al. [65,66] and Iwaoka et al. [67], respectively. (c) Oxidative folding pathways of αLA. The red and blue arrows are major pathways in the presence and absence of Ca2+, respectively.

Molecules 2021, 26, 195 12 of 19

This classical scenario has been revised recently by applying DHSox as an oxidant [67]. When R was reacted with stoichiometric amounts of DHSox, not only 1SS and 2SS but also 3SS and 4SS were generated in a similar manner that was observed for hirudin and RNase A (see above). Then, the 3SS was gradually converted to the three des intermediates via slow SS rearrangement (Figure8b). More importantly, des[64-80] and des[6-127], which were located on the major folding pathways at 25 ◦C and below, were not formed at 35 ◦C. Instead, only des[76-94] was observed at the temperature, and it was directly oxidized to N by addition of DHSox. Thus, the major folding pathways at the physiological temperature can be invisible at the low temperatures. Bovine α-lactalbumin (αLA) is another representative four-disulﬁde protein, for which the oxidative folding pathways were well studied [68]. Although αLA shares a similar native structure as well as a SS topology with HEL, the oxidative folding pathways were essentially different (Figure8c). Since αLA had a calcium binding pocket in the β-sheet domain, in which two SS linkages of Cys61–Cys77 and Cys73–Cys91 are located, the presence of a calcium ion (Ca2+) greatly affected the oxidative folding pathways as well as the folding velocity and the yield of native αLA (N) [69–71]. Indeed, under a metal- free condition, SS formation of reduced αLA (R) took place at random and produced heterogeneous SS intermediate ensembles (1SS, 2SS, 3SS, and 4SS), resulting in a low yield of N (a hirudin-like folding). In the presence of Ca2+, on the other hand, two structured SS intermediates, which were assigned to (61-77, 73-91) and des[6-120], were rapidly populated (a BPTI-like folding). The (61-77, 73-91) intermediate would form by SS rearrangement of 2SS and can subsequently be oxidized to N through the des[6-120] intermediate. Thus, the oxidative folding pathways of αLA can be controlled by the presence of a metal ion. In this case, however, the folding pathways were robust against the temperature change [71].

4.4. Oxidative Folding of Proteins with Odd Cys Residues With few exceptions, many small globular proteins have an even number of cysteine residues and no free thiol group in the native state. Therefore, the oxidative folding pathways of a protein with odd Cys residues were not well known until recently. Probably the first extensive study of such a case was reported using bovine milk β-lactoglobulin (BLG) as a model protein [72]. BLG is a major whey protein with one free cysteinyl thiol (Cys121) in addition to two SS bonds (Cys66–Cys160 and Cys106–Cys119). While the SS-intact conformational folding had been studied in depth [73], the oxidative folding pathways remained unknown. When reduced BLG variant A (BLGA) (R) was oxidized with DHSox, a limited number of SS intermediates were observed (Figure9). Among the produced SS intermediates, two distinct 1SS intermediates, which have native Cys66–Cys160 SS bond (I-1) or another native Cys106–Cys119 SS bond (I-2), could be characterized. In the major folding pathway, R, which was rich in α-helices, was firstly oxidized to I-1 and subsequently converted to I-2 via SS rearrangement. Finally, I-2 gained the missing Cys66–Cys160 SS bond by oxidation to regenerate native BLGA (N), which is rich in β-sheets. In addition to this major folding pathway, N could also be regenerated from the scrambled 2SS intermediate ensemble via SS rearrangement, but this minor pathway would be deteriorative because the 2SS was prone to self-aggregate irreversibly probably due to the presence of a free Cys SH group. During the oxidative folding of BLGA, the redundant Cys121 SH group may assist efficient SS-rearrangement of the SS intermediates, guiding them to I-1 and then I-2, not to 2SS [72]. The oxidative folding pathways of odd Cys proteins remain largely unknown and will be interesting targets in the future study. Molecules 2021, 26, x FOR PEER REVIEW 13 of 20

Cys106–Cys119 SS bond (I-2), could be characterized. In the major folding pathway, R, which was rich in α-helices, was firstly oxidized to I-1 and subsequently converted to I-2 via SS rearrangement. Finally, I-2 gained the missing Cys66–Cys160 SS bond by oxidation to regenerate native BLGA (N), which is rich in β-sheets. In addition to this major folding pathway, N could also be regenerated from the scrambled 2SS intermediate ensemble via SS rearrangement, but this minor pathway would be deteriorative because the 2SS was prone to self-aggregate irreversibly probably due to the presence of a free Cys SH group. During the oxidative folding of BLGA, the redundant Cys121 SH group may assist efficient SS-rearrangement of the SS intermediates, guiding them to I-1 and then I-2, not to 2SS [72]. Molecules 2021, 26, 195 The oxidative folding pathways of odd Cys proteins remain largely unknown13 ofand 19 will be interesting targets in the future study.

Molecules 2021, 26, x FOR PEER REVIEW 14 of 20

tions between reduced and non-protected bovine Ins-A and -B chains under various conditions, which included the conditions using DHSox as an oxidant [74]. The revealed major FigureFigure 9. 9.Oxidative Oxidative folding foldingfolding pathways pathways pathways of of bovine bovine are milk milk summarizedβ β-lactoglobulin-lactoglobulin in (BLGA) Figure(BLGA) with 11,with where odd odd Cys Cy 1SSs residues. residues.A and 2SS Red RedA and representand black black arrows arrows one- and representrepresent major major and and minor minortwo-SS pathways, pathways, intermediate respectively. respectively. ensembles, respectively, of Ins-A, the major component of which has a native CysA6–CysA11 SS bond. In the folding pathways, 2SS*, a heterodimeric 2SS isomer having4.5.4.5. OxidativeOxidative two out FoldingFolding of three PathwaysPathways native SS ofof Twobonds,Two Chain Chain i.e., Proteins ProteinsCysA6–Cys A11 and CysA20–CysB19, is a pivotal precursorElucidationElucidation of N. The of of oxidative spectroscopicoxidative folding folding characterization pathways pathways of of two demonstratedtwo chain chain proteins proteins that is another2SS*is another has interesting a native-interest- likeanding folded and challenging challenging structure, issue. issue.indicating With With this that this regard, theregard, 2SS* the the preferentially folding folding pathway pathway converts of of insulin insulinto N, wasotherwise was recently recently it mightcharacterizedcharacterized easily dissociate ﬁnally finally and and to has2SS hasA been andbeen appliedR appliedB through to to facile intramolecular facile chemical chemical synthesis SS synthesis reshuffling. of insulinof insulin The by key aby native 2SS* a na- intermediatechaintive chain assembly assembly would (NCA) be (NCA) generated method method [74 through]. [74]. several paths, such as from 1SSA by the reaction with bothInsulinInsulin RB isandis aa smallan oxidant, globular from protein protein 1SS*, (5.8which (5.8 kDa), kDa), has comprised compriseda native SS of ofcrosslink two two peptide peptide between chains, chains, Cys i.e.,A20 i.e., A-– CysA-chainchainB19, by(Ins-A, (Ins-A, oxidation, 21 21 amino amino and acid from acid residues) residues)2SSA through and and B-ch B-chain the ainintermolecular (Ins-B, 30 amino-acid SS-SH exchange residues),residues), reaction whichwhich withareare crosslinkedcrosslinkedRB. The revealed by by Cys Cys two-chainA7A7-Cys-CysB7B7 and andfolding CysCys A20pathwaysA20-Cys-CysB19 SSof insulin bridges strongly (Figure(Figure 10suggest10a).a). AnotherAnother that the SSSS yieldbond,bond, of Cys CysinsulinA6A6-Cys-Cys byA11A11 NCA,, is is located locatedcan be increasedin the A-chain. by stabilizing InIn pancreaticpancreatic 2SS* byββ-cells,-cells, any methods. insulininsulin isis Predictably, producedproduced as as underproinsulin,proinsulin, the optimized a a single single polypeptide polypeptideNCA condition chain chain of(at of 86 pH 86 amino amino10.0 acidand acid residues,−10 residues, °C), bovine in whichin which pancreatic the the C terminus C terminus insulin of (BPIns)theof the B-chain B-chainand ishuman linked is linked insulin to the to N-terminus the(HIns) N-terminus were of theobtained of A-chain the A-chainin byup a to connecting by 39% a connecting and C-peptide. 49% yields, C-peptide. The respec- proin- The tively.sulinproinsulin is Similarly, effectively is effectively human oxidized type-II oxidized in ER relaxin to in obtain ER (Rlx2) to the obtain (Figure10b), three the native three SSwhich bondsnative is andaSS pregnancy bonds is then and converted peptide is then hormonetoconverted the mature and to a the two-chainrepresentative mature insulintwo-chain member by proteolyticinsulin of the byinsulin proteolytic cleavage family, of cleavagewas the C-peptide.regenerated of the C-peptide. The in up oxidative to 49% The byfoldingoxidative applying pathways folding the NCA ofpathways proinsulin condition of in[74].proinsulin vivo were in well vivo characterized were well characterized by Feng et al. by [75 Feng,76]. et al. [75,76]. There is a long history of the attempts to artificially prepare insulin and its analogues. The initial attempt was a very simple approach of just mixing the reduced A- and B-chains in a folding buffer solution [77–79]. However, this primitive strategy disappointingly resulted in a low yield of insulin (less than 5%). Following this unsuccessful attempt, a variety of attempts, such as an approach of the directed SS crosslinking between the two chains and a biomimetic approach of the proinsulin-like single-chain folding, have been considered to increase the insulin yield. However, none of them has been fabricated to the practical level to supply insulin in the market. In this context, to know the exact two-chain folding pathways of insulin would be a sober but steady approach. According to this line, we carried out the crosslinking reac-

FigureFigure 10. 10. 3D3D structures structures of of insulin insulin and and relaxin. relaxin. ( (aa)) Bovine Bovine pancreatic pancreatic insulin insulin (BPIns). (BPIns). ( (bb)) Human Human relaxin relaxin 2 2 (Rlx). (Rlx). There is a long history of the attempts to artiﬁcially prepare insulin and its analogues. The initial attempt was a very simple approach of just mixing the reduced A- and B-chains in a folding buffer solution [77–79]. However, this primitive strategy disappointingly resulted in a low yield of insulin (less than 5%). Following this unsuccessful attempt, a variety of attempts, such as an approach of the directed SS crosslinking between the two chains and a biomimetic approach of the proinsulin-like single-chain folding, have been Molecules 2021, 26, 195 14 of 19

considered to increase the insulin yield. However, none of them has been fabricated to the practical level to supply insulin in the market. In this context, to know the exact two-chain folding pathways of insulin would be a sober but steady approach. According to this line, we carried out the crosslinking reactions between reduced and non-protected bovine Ins-A and -B chains under various conditions, which included the conditions using DHSox as an oxidant [74]. The revealed major folding pathways are summarized in Figure 11, where 1SSA and 2SSA represent one- and two-SS intermediate ensembles, respectively, of Ins-A, the major component of which has a native CysA6–CysA11 SS bond. In the folding pathways, 2SS*, a heterodimeric 2SS isomer having two out of three native SS bonds, i.e., CysA6–CysA11 and CysA20–CysB19, is a pivotal precursor of N. The spectroscopic characterization demonstrated that 2SS* has a native-like folded structure, indicating that the 2SS* preferentially converts to N, otherwise it might easily dissociate to 2SSA and RB through intramolecular SS reshufﬂing. The key 2SS* intermediate would be generated through several paths, such as from 1SSA by the reaction with both RB and an oxidant, from 1SS*, which has a native SS crosslink between CysA20–CysB19, by oxidation, and from 2SSA through the intermolecular SS-SH exchange reaction with RB. The revealed two-chain folding pathways of insulin strongly suggest that the yield of insulin by NCA can be increased by stabilizing 2SS* by any methods. Predictably, under the optimized NCA condition (at pH 10.0 and −10 ◦C), bovine pancreatic insulin (BPIns) and human insulin (HIns) were obtained in up to 39% and 49% yields, respectively. Similarly, human type-II relaxin (Rlx2) (Figure 10b), which is Molecules 2021, 26, x FOR PEER REVIEWa pregnancy peptide hormone and a representative member of the insulin family,15 was of 20 regenerated in up to 49% by applying the NCA condition [74].

Number of SS linkage

0 1 2 3

Unfolded RA

6 SH Metastable intermediate

Conformation 11 ox 7HS SH 2SS* HS 20 S red 7 19S Slow! SS SS

Folded N (BPIns)

FigureFigure 11. 11.Oxidative Oxidative two-chain two-chain folding folding pathways pathways of of BPIns BPIns revealed revealed by by the the native native chain chain assembly assembly method. method. Application of the SS-to-SeSe strategy (see Sections 4.1 and 4.2) to insulin folding has Application of the SS-to-SeSe strategy (see Sections 4.1 and 4.2) to insulin folding has already been applied by mixing the synthesized Sec-replaced Ins-A and -B chains, i.e., C7UA already been applied by mixing the synthesized Sec-replaced Ins-A and -B chains, i.e., and C7UB, under the NCA condition [80]. The [C7UA, C7UB] insulin analogue, so-called C7UA and C7UB, under the NCA condition [80]. The [C7UA, C7UB] insulin analogue, so- selenoinsulin (Se-Ins) (Figure 12), was efﬁciently generated in up to 27% isolated yield even called selenoinsulin (Se-Ins) (Figure 12), was efficiently generated in up to 27% isolated under slightly milder condition (at 4 ◦C) than that employed for NCA of wild-type insulin. yield even under slightly milder condition (at 4 °C) than that employed for NCA of wild- type insulin. The folding of Se-Ins would be assisted by the spontaneous chain assembly probably via selective formation of the SecA7-SecB7 SeSe crosslink. Thus, the SS-to-SeSe strategy can be an efficient method to modulate the folding pathways of two-chain proteins too. Metanis and coworkers [81] have also performed a similar chain assembly synthesis of human insulin analogue, i.e., a [C6UA, C11UA] Se-Ins variant, by reacting the Sec-replaced A-chain with the wild-type B-chain. Both the Se-Ins’s prepared by us [80] and Me- tanis [81] exhibited a bioactivity comparable to the wild-type insulin, indicating that the molecular structures of Se-Ins are almost the same as the wild-type in solution.

Figure 12. Primary amino acid sequences of the selenoinsulin series. (a) Bovine selenoinsulin reported by Arai et al. [80]. (b) Human selenoinsulin reported by Metanis et al. [81].

5. Flexibility of the Oxidative Folding Pathways During the oxidation of a polypeptide chain having several Cys residues, two ele- mentary reactions, i.e., SS formation and SS rearrangement, are necessarily involved (Fig- ure 2). These reactions indeed control the oxidative folding pathways to the native state. Molecules 2021, 26, x FOR PEER REVIEW 15 of 20

Number of SS linkage

0 1 2 3

Unfolded RA

6 SH Metastable intermediate

Conformation 11 ox 7HS SH 2SS* HS 20 S red 7 19S Slow! SS SS

Folded N (BPIns)

Figure 11. Oxidative two-chain folding pathways of BPIns revealed by the native chain assembly method.

Application of the SS-to-SeSe strategy (see Sections 4.1 and 4.2) to insulin folding has already been applied by mixing the synthesized Sec-replaced Ins-A and -B chains, i.e., C7UA and C7UB, under the NCA condition [80]. The [C7UA, C7UB] insulin analogue, so- called selenoinsulin (Se-Ins) (Figure 12), was efficiently generated in up to 27% isolated yield even under slightly milder condition (at 4 °C) than that employed for NCA of wild- type insulin. The folding of Se-Ins would be assisted by the spontaneous chain assembly probably via selective formation of the SecA7-SecB7 SeSe crosslink. Thus, the SS-to-SeSe strategy can be an efficient method to modulate the folding pathways of two-chain pro- Molecules 2021, 26, 195 teins too. 15 of 19 Metanis and coworkers [81] have also performed a similar chain assembly synthesis of human insulin analogue, i.e., a [C6UA, C11UA] Se-Ins variant, by reacting the Sec-re- The folding of Se-Ins would be assisted by the spontaneous chain assembly probably via placed A-chain with the wild-type B-chain. Both the Se-Ins’s prepared by us [80] and Me- selective formation of the SecA7-SecB7 SeSe crosslink. Thus, the SS-to-SeSe strategy can be tanis [81] exhibited a bioactivity comparable to the wild-type insulin, indicating that the an efﬁcient method to modulate the folding pathways of two-chain proteins too. molecular structures of Se-Ins are almost the same as the wild-type in solution.

Figure 12. Primary amino acid sequences of the selenoinsulin series. (a) Bovine selenoinsulin reported by Arai et al. [80]. Figure 12. Primary amino acid sequences of the selenoinsulin series. (a) Bovine selenoinsulin reported by Arai et al. [80]. (b) Human selenoinsulin reported by Metanis et al. [81]. (b) Human selenoinsulin reported by Metanis et al. [81].

5. FlexibilityMetanis andof the coworkers Oxidative [81 Folding] have also Pathways performed a similar chain assembly synthesis of humanDuring insulin the analogue, oxidation i.e., of aa [C6UpolypeptideA, C11U Ach] Se-Insain having variant, several by reacting Cys residues, the Sec-replaced two ele- mentaryA-chain withreactions, the wild-type i.e., SS formation B-chain. Bothand SS the rearrangement, Se-Ins’s prepared are by necessarily us [80] and involved Metanis (Fig- [81] ureexhibited 2). These a bioactivity reactions comparableindeed control to thethe wild-typeoxidative insulin,folding indicatingpathways to that the the native molecular state. structures of Se-Ins are almost the same as the wild-type in solution.

5. Flexibility of the Oxidative Folding Pathways During the oxidation of a polypeptide chain having several Cys residues, two el- ementary reactions, i.e., SS formation and SS rearrangement, are necessarily involved (Figure2) . These reactions indeed control the oxidative folding pathways to the native state. As seen above for several model peptides and proteins, there seems to be no general rule how to delineate the oxidative folding pathways based on the number and topology of the SS bonds present in the native folded state. This feature is obvious by looking at the significantly different pathways of HEL and αLA (Figure8), which are classified to the same protein superfamily [82]. Thus, oxidative folding pathways of a protein would depend primarily on the amino acid sequence. This in turn suggests that the information about the folding pathways, not only the 3D folded structure, of a protein would also be encoded in the amino acid sequence. In the previous stage of protein folding study, oxidative folding pathways were clearly characterized for several representative SS-containing proteins. However, as exemplified above, recent reinvestigation of those pathways applying a new oxidant as well as an artificially designed polypeptide, has disclosed that the folding pathways are significantly flexible depending on the conditions, such as the temperature, pH, and co-existing metal ions. In some cases, the previously proposed pathways may be hindered, or the previously hidden pathways may be possible, under physiologically relevant conditions. Indeed, in the synthesis of a protein in vivo, the polypeptide chain elongated on the ribosome is trans- ported into the endoplasmic reticulum (ER), where the oxidative folding is initiated from the N-terminal domain in the presence of various SS-oxidoreductases and SS-isomerases, such as PDI and its family proteins [83]. Thus, the oxidative folding pathway of a full-length polypeptide characterized in a test tube could be different from that in the ER. The roles of PDI family proteins are especially important because they efficiently promote the oxidative folding by retrieving kinetically trapped intermediates back to the normal pathways and also by avoiding the formation of metastable oligomers or aggregates. This indicates that ER-resident PDI family enzymes and possibly their mimic catalysts as well would be able to control the major oxidative folding pathways in vitro. Thus, an appropriate use of the enzymes and mimics should be a practical strategy to improve the yield and velocity of the oxidative folding that is often a bottleneck process in chemical synthesis of proteins [26–35]. From the viewpoint of peptide and protein engineering, flexibility of the folding pathways would be fascinating. It would be applicable to control the structures, thereby Molecules 2021, 26, 195 16 of 19

functions, of proteins by changing the folding conditions. For example, by changing folding conditions on purpose, a polypeptide chain might fold to a new structure through a different pathway. More practically, by applying the recently available new folding strategies, such as a use of a strong and selective oxidant and the substitution of Cys by Sec, the proteins that could not be folded under classical folding conditions would be obtained effectively. Prevention and treatment of protein misfolding diseases would also be possible by controlling the protein folding pathways under well-managed folding conditions.

6. Perspectives and Concluding Remarks In the classical oxidative folding study, the folding reaction was frequently conducted under non-natural environments so that the key SS intermediates could be easily observed. This indeed allowed the researchers to characterize the oxidative folding pathways of some representative proteins and propose a general model of protein folding (Figure1). At the same time, they gradually noticed that the folding pathways greatly depends on the reaction environments as well as the type of proteins. In this review, recent achievements in the relevant fields are overviewed. It should be emphasized here that the folding pathways of peptides and proteins are significantly flexible. In relation to prevention and treatment of protein misfolding diseases, it is preferable to investigate the protein folding pathways under the conditions that are close to physiological environments, if possible, at an organ level, where numerous factors coexist. In the meantime, the methods to control the folding pathways are now available by applying the state-of-the-art technologies as illustrated in this review. Engineering peptide and protein structures and the functions by controlling the folding pathways are interesting and important issues in the new realm of the current protein folding study.

Author Contributions: K.A. and M.I. collaboratively wrote this review article. These authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References 1. Anfinsen, C.B. The formation and stabilization of protein structure. Biochem. J. 1972, 128, 737–749. [CrossRef] 2. Anfinsen, C.B. Principles that Govern the Folding of Protein Chains. Science 1973, 181, 223–230. [CrossRef] 3. Levinthal, C. Are there pathways for protein folding? J. Chim. Phys. Phys.-Chim. Biol. 1968, 65, 44–45. [CrossRef] 4. Arolas, J.L.; Aviles, F.X.; Chang, J.-Y.; Ventura, S. Folding of small disulfide-rich proteins: Clarifying the puzzle. Trends Biochem. Sci. 2006, 31, 292–301. [CrossRef][PubMed] 5. Edsall, J.T. Hsien Wu and the First Theory of Protein Denaturation (1931). In Advances in Protein Chemistry; Anfinsen, C.B., Richards, F.M., Edsall, J.T., Eisenberg, D.S., Eds.; Protein Stability; Academic Press: Cambridge, MA, USA, 1995; Volume 46, pp. 1–5. [CrossRef] 6. Mossuto, M.F. Disulfide Bonding in Neurodegenerative Misfolding Diseases. Int. J. Cell Biol. 2013, 2013, 318319. [CrossRef] [PubMed] 7. Hartl, F.U. Protein Misfolding Diseases. Annu. Rev. Biochem. 2017, 86, 21–26. [CrossRef][PubMed] 8. Khoury, G.A.; Smadbeck, J.; Kieslich, C.A.; Floudas, C.A. Protein folding and de novo protein design for biotechnological applications. Trends Biotechnol. 2014, 32, 99–109. [CrossRef] 9. Rigoldi, F.; Donini, S.; Redaelli, A.; Parisini, E.; Gautieri, A. Review: Engineering of thermostable enzymes for industrial applications. APL Bioeng. 2018, 2.[CrossRef] 10. Englander, S.W.; Mayne, L. The nature of protein folding pathways. Proc. Natl. Acad. Sci. USA 2014, 111, 15873–15880. [CrossRef] 11. Bulaj, G.; Kortemme, T.; Goldenberg, D.P. Ionization−Reactivity Relationships for Cysteine Thiols in Polypeptides. Biochemistry 1998, 37, 8965–8972. [CrossRef] 12. Tajc, S.G.; Tolbert, B.S.; Basavappa, R.; Miller, B.L. Direct Determination of Thiol pKa by Isothermal Titration Microcalorimetry. J. Am. Chem. Soc. 2004, 126, 10508–10509. [CrossRef][PubMed] 13. Aitken, A.; Learmonth, M. Carboxymethylation of Cysteine Using Iodoacetamide/ Iodoacetic Acid. In The Protein Protocols Handbook; Springer Protocols Handbooks; Walker, J.M., Ed.; Humana Press: Totowa, NJ, USA, 2002; pp. 455–456. ISBN 978-1- 59259-169-5. 14. Narayan, M.; Welker, E.; Wedemeyer, W.J.; Scheraga, H.A. Oxidative Folding of Proteins. Acc. Chem. Res. 2000, 33, 805–812. [CrossRef][PubMed] Molecules 2021, 26, 195 17 of 19

15. Wolynes, P.G.; Onuchic, J.N.; Thirumalai, D. Navigating the folding routes. Science 1995, 267, 1619–1620. [CrossRef][PubMed] 16. Narayan, M. Revisiting the Formation of a Native Disulfide Bond: Consequences for Protein Regeneration and Beyond. Molecules 2020, 25, 5337. [CrossRef][PubMed] 17. Rothwarf, D.M.; Scheraga, H.A. Regeneration of bovine pancreatic ribonuclease A. 3. Dependence on the nature of the redox reagent. Biochemistry 1993, 32, 2690–2697. [CrossRef][PubMed] 18. Beld, J.; Woycechowsky, K.J.; Hilvert, D. Selenoglutathione: Efficient Oxidative Protein Folding by a Diselenide. Biochemistry 2007, 46, 5382–5390. [CrossRef] 19. Arai, K.; Dedachi, K.; Iwaoka, M. Rapid and Quantitative Disulfide Bond Formation for a Polypeptide Chain Using a Cyclic Selenoxide Reagent in an Aqueous Medium. Chem. Eur. J. 2011, 17, 481–485. [CrossRef] 20. Arai, K.; Iwaoka, M. Oxidative Protein Folding Using trans-3,4-Dihydroxyselenolane Oxide, Chapter 14. In Functional Disulphide Bonds: Methods and Protocols; Methods in Molecular Biology; Hogg, P., Ed.; Humana Press: Totowa, NJ, USA, 2019; ISBN 978-1- 59259-169-5. 21. Arai, K.; Noguchi, M.; Singh, B.G.; Priyadarsini, K.I.; Fujio, K.; Kubo, Y.; Takayama, K.; Ando, S.; Iwaoka, M. A water-soluble selenoxide reagent as a useful probe for the reactivity and folding of polythiol peptides. FEBS Open Bio 2013, 3, 55–64. [CrossRef] 22. Iwaoka, M.; Kumakura, F.; Yoneda, M.; Nakahara, T.; Henmi, K.; Aonuma, H.; Nakatani, H.; Tomoda, S. Direct Observation of Conformational Folding Coupled with Disulphide Rearrangement by Using a Water-soluble Selenoxide Reagent—A Case of Oxidative Regeneration of Ribonuclease A under Weakly Basic Conditions. J. Biochem. 2008, 144, 121–130. [CrossRef] 23. Shi, T.; Rabenstein, D.L. Discovery of a Highly Selective and Efficient Reagent for Formation of Intramolecular Disulfide Bonds in Peptides. J. Am. Chem. Soc. 2000, 122, 6809–6815. [CrossRef] 24. Narayan, M.; Welker, E.; Scheraga, H.A. Native Conformational Tendencies in Unfolded Polypeptides: Development of a Novel Method To Assess Native Conformational Tendencies in the Reduced Forms of Multiple Disulfide-Bonded Proteins. J. Am. Chem. Soc. 2003, 125, 2036–2037. [CrossRef] 25. Narayan, M.; Welker, E.; Wanjalla, C.; Xu, G.; Scheraga, H.A. Shifting the Competition between the Intramolecular Reshuffling Reaction and the Direct Oxidation Reaction during the Oxidative Folding of Kinetically Trapped Disulfide-Insecure Intermediates. Biochemistry 2003, 42, 10783–10789. [CrossRef] 26. Lees, W.J. Small-molecule catalysts of oxidative protein folding. Curr. Opin. Chem. Biol. 2008, 12, 740–745. [CrossRef][PubMed] 27. Madar, D.J.; Patel, A.S.; Lees, W.J. Comparison of the oxidative folding of lysozyme at a high protein concentration using aromatic thiols versus glutathione. J. Biotechnol. 2009, 142, 214–219. [CrossRef][PubMed] 28. Potempa, M.; Hafner, M.; Frech, C. Mechanism of Gemini Disulfide Detergent Mediated Oxidative Refolding of Lysozyme in a New Artificial Chaperone System. Protein J. 2010, 29, 457–465. [CrossRef] 29. Metanis, N.; Foletti, C.; Beld, J.; Hilvert, D. Selenoglutathione-Mediated Rescue of Kinetically Trapped Intermediates in Oxidative Protein Folding. Isr. J. Chem. 2011, 51, 953–959. [CrossRef] 30. Iii, J.C.L.; Andersen, K.A.; Wallin, K.K.; Raines, R.T. Organocatalysts of oxidative protein folding inspired by protein disulfide isomerase. Org. Biomol. Chem. 2014, 12, 8598–8602. [CrossRef] 31. Reddy, P.S.; Metanis, N. Small molecule diselenide additives for in vitro oxidative protein folding. Chem. Commun. 2016, 52, 3336–3339. [CrossRef] 32. Arai, K.; Ueno, H.; Asano, Y.; Chakrabarty, G.; Shimodaira, S.; Mugesh, G.; Iwaoka, M. Protein Folding in the Presence of Water-Soluble Cyclic Diselenides with Novel Oxidoreductase and Isomerase Activities. ChemBioChem 2018, 19, 207–211. [CrossRef] 33. Okada, S.; Matsusaki, M.; Arai, K.; Hidaka, Y.; Inaba, K.; Okumura, M.; Muraoka, T. Coupling effects of thiol and urea-type groups for promotion of oxidative protein folding. Chem. Commun. 2019, 55, 759–762. [CrossRef] 34. Tsukagoshi, S.; Mikami, R.; Arai, K. Basic Amino Acid Conjugates of 1,2-Diselenan-4-amine with Protein Disulfide Isomerase-like Functions as a Manipulator of Protein Quality Control. Chem. Asian J. 2020, 15, 2646–2652. [CrossRef] 35. Kersteen, E.A.; Raines, R.T. Catalysis of Protein Folding by Protein Disulfide Isomerase and Small-Molecule Mimics. Antioxid. Redox Signal. 2003, 5, 413–424. [CrossRef] 36. Pegoraro, S.; Fiori, S.; Rudolph-Böhner, S.; Watanabe, T.X.; Moroder, L. Isomorphous replacement of cystine with selenocystine in endothelin: Oxidative refolding, biological and conformational properties of [Sec3,Sec11,Nle7]-endothelin-11 1Edited by R. Huber. J. Mol. Biol. 1998, 284, 779–792. [CrossRef][PubMed] 37. Iwaoka, M.; Arai, K. From Sulfur to Selenium. A New Research Arena in Chemical Biology and Biological Chemistry. Curr. Chem. Biol. 2013, 7, 2–24. [CrossRef] 38. Pegoraro, S.; Fiori, S.; Cramer, J.; Rudolph-Böhner, S.; Moroder, L. The disulfide-coupled folding pathway of apamin as derived from diselenide-quenched analogs and intermediates. Protein Sci. 1999, 8, 1605–1613. [CrossRef][PubMed] 39. Fiori, S.; Pegoraro, S.; Rudolph-Böhner, S.; Cramer, J.; Moroder, L. Synthesis and conformational analysis of apamin analogues with natural and non-natural cystine/selenocystine connectivities. Biopolymers 2000, 53, 550–564. [CrossRef] 40. Moroder, L.; Musiol, H.-J.; Götz, M.; Renner, C. Synthesis of single- and multiple-stranded cystine-rich peptides. Pept. Sci. 2005, 80, 85–97. [CrossRef] 41. Walewska, A.; Zhang, M.-M.; Skalicky, J.J.; Yoshikami, D.; Olivera, B.M.; Bulaj, G. Integrated Oxidative Folding of Cys- teine/Selenocysteine Containing Peptides: Improving Chemical Synthesis of Conotoxins. Angew. Chem. Int. Ed. 2009, 48, 2221–2224. [CrossRef] Molecules 2021, 26, 195 18 of 19

42. Han, T.S.; Zhang, M.-M.; Gowd, K.H.; Walewska, A.; Yoshikami, D.; Olivera, B.M.; Bulaj, G. Disulfide-Depleted Selenoconopep- tides: Simplified Oxidative Folding of Cysteine-Rich Peptides. ACS Med. Chem. Lett. 2010, 1, 140–144. [CrossRef] 43. de Araujo, A.D.; Callaghan, B.; Nevin, S.T.; Daly, N.L.; Craik, D.J.; Moretta, M.; Hopping, G.; Christie, M.J.; Adams, D.J.; Alewood, P.F. Total Synthesis of the Analgesic Conotoxin MrVIB through Selenocysteine-Assisted Folding. Angew. Chem. Int. Ed. 2011, 50, 6527–6529. [CrossRef] 44. Walewska, A.; Ja´skiewicz,A.; Bulaj, G.; Rolka, K. Selenopeptide Analogs of EETI-II Retain Potent Trypsin Inhibitory Activities. Chem. Biol. Drug Des. 2011, 77, 93–97. [CrossRef][PubMed] 45. Steiner, A.M.; Woycechowsky, K.J.; Olivera, B.M.; Bulaj, G. Reagentless Oxidative Folding of Disulfide-Rich Peptides Catalyzed by an Intramolecular Diselenide. Angew. Chem. Int. Ed. 2012, 51, 5580–5584. [CrossRef][PubMed] 46. Creighton, T.E. Experimental studies of protein folding and unfolding. Prog. Biophys. Mol. Biol. 1979, 33, 231–297. [CrossRef] 47. Creighton, T.E.; Goldenberg, D.P. Kinetic role of a meta-stable native-like two-disulphide species in the folding transition of bovine pancreatic trypsin inhibitor. J. Mol. Biol. 1984, 179, 497–526. [CrossRef] 48. Weissman, J.S.; Kim, P.S. Reexamination of the folding of BPTI: Predominance of native intermediates. Science 1991, 253, 1386–1393. [CrossRef] 49. Weissman, J.S.; Kim, P.S. Kinetic role of nonnative species in the folding of bovine pancreatic trypsin inhibitor. Proc. Natl. Acad. Sci. USA 1992, 89, 9900–9904. [CrossRef] 50. Weissman, J.S.; Kim, P.S. A kinetic explanation for the rearrangement pathway of BPTI folding. Nat. Struct. Biol. 1995, 2, 1123–1130. [CrossRef] 51. Creighton, T.E.; Darby, N.J.; Kemmink, J. The roles of partly folded intermediates in protein folding. FASEB J. 1996, 10, 110–118. [CrossRef] 52. Metanis, N.; Hilvert, D. Strategic Use of Non-Native Diselenide Bridges to Steer Oxidative Protein Folding. Angew. Chem. Int. Ed. 2012, 51, 5585–5588. [CrossRef] 53. Metanis, N.; Hilvert, D. Harnessing selenocysteine reactivity for oxidative protein folding. Chem. Sci. 2014, 6, 322–325. [CrossRef] 54. Mousa, R.; Lansky, S.; Shoham, G.; Metanis, N. BPTI folding revisited: Switching a disulfide into methylene thioacetal reveals a previously hidden path. Chem. Sci. 2018, 9, 4814–4820. [CrossRef] 55. Chatrenet, B.; Chang, J.Y. The folding of hirudin adopts a mechanism of trial and error. J. Biol. Chem. 1992, 267, 3038–3043. [PubMed] 56. Chatrenet, B.; Chang, J.Y. The disulfide folding pathway of hirudin elucidated by stop/go folding experiments. J. Biol. Chem. 1993, 268, 20988–20996. 57. Chang, J.-Y. The Properties of Scrambled Hirudins. J. Biol. Chem. 1995, 270, 25661–25666. [CrossRef][PubMed] 58. Thannhauser, T.W.; Rothwarf, D.M.; Scheraga, H.A. Kinetic Studies of the Regeneration of Recombinant Hirudin Variant 1 with Oxidized and Reduced Dithiothreitol. Biochemistry 1997, 36, 2154–2165. [CrossRef][PubMed] 59. Komatsu, Y.; Misawa, S.; Sukesada, A.; Ohba, Y.; Hayashi, H. CX-397, a Novel Recombinant Hirudin Analog Having a Hybrid Sequence of Hirudin Variants-1 and -3. Biochem. Biophys. Res. Commun. 1993, 196, 773–779. [CrossRef][PubMed] 60. Shimamoto, S.; Fukutsuji, M.; Osumi, T.; Goto, M.; Toyoda, H.; Hidaka, Y. Topological Regulation of the Bioactive Conformation of a Disulfide-Rich Peptide, Heat-Stable Enterotoxin. Molecules 2020, 25, 4798. [CrossRef] 61. Rothwarf, D.M.; Li, Y.-J.; Scheraga, H.A. Regeneration of Bovine Pancreatic Ribonuclease A: Identification of Two Nativelike Three-Disulfide Intermediates Involved in Separate Pathways. Biochemistry 1998, 37, 3760–3766. [CrossRef] 62. Welker, E.; Narayan, M.; Volles, M.J.; Scheraga, H.A. Two new structured intermediates in the oxidative folding of RNase A. FEBS Lett. 1999, 460, 477–479. [CrossRef] 63. Arai, K.; Kumakura, F.; Iwaoka, M. Characterization of Kinetic and Thermodynamic Phases in the Prefolding Process of Bovine Pancreatic Ribonuclease A Coupled with Fast SS Formation and SS Reshuffling. Biochemistry 2010, 49, 10535–10542. [CrossRef] 64. Arai, K.; Kumakura, F.; Iwaoka, M. Kinetic and thermodynamic analysis of the conformational folding process of SS-reduced bovine pancreatic ribonuclease A using a selenoxide reagent with high oxidizing ability. FEBS Open Bio 2012, 2, 60–70. [CrossRef] [PubMed] 65. van den Berg, B.; Chung, E.W.; Robinson, C.V.; Mateo, P.L.; Dobson, C.M. The oxidative refolding of hen lysozyme and its catalysis by protein disulfide isomerase. EMBO J. 1999, 18, 4794–4803. [CrossRef] 66. van den Berg, B.; Chung, E.W.; Robinson, C.V.; Dobson, C.M. Characterisation of the dominant oxidative folding intermediate of hen lysozyme. J. Mol. Biol. 1999, 290, 781–796. [CrossRef] 67. Arai, K.; Shibagaki, W.; Shinozaki, R.; Iwaoka, M. Reinvestigation of the Oxidative Folding Pathways of Hen Egg White Lysozyme: Switching of the Major Pathways by Temperature Control. Int. J. Mol. Sci. 2013, 14, 13194–13212. [CrossRef] 68. Ewbank, J.J.; Creighton, T.E. Structural characterization of the disulfide folding intermediates of bovine alpha-lactalbumin. Biochemistry 1993, 32, 3694–3707. [CrossRef] 69. Chang, J.-Y. The Folding Pathway of α-Lactalbumin Elucidated by the Technique of Disulfide Scrambling. Isolation of on-pathway and off-pathway intermediates. J. Biol. Chem. 2002, 277, 120–126. [CrossRef][PubMed] 70. Chang, J.-Y.; Li, L. Pathway of Oxidative Folding of α-Lactalbumin: A Model for Illustrating the Diversity of Disulfide Folding Pathways. Biochemistry 2002, 41, 8405–8413. [CrossRef][PubMed] 71. Shinozaki, R.; Iwaoka, M. Effects of Metal Ions, Temperature, and a Denaturant on the Oxidative Folding Pathways of Bovine α-Lactalbumin. Int. J. Mol. Sci. 2017, 18, 1996. [CrossRef][PubMed] Molecules 2021, 26, 195 19 of 19

72. Iwaoka, M.; Mitsuji, T.; Shinozaki, R. Oxidative folding pathways of bovine milk β-lactoglobulin with odd cysteine residues. FEBS Open Bio 2019, 9, 1379–1391. [CrossRef] 73. Ikeguchi, M. Transient Non-Native Helix Formation during the Folding of β-Lactoglobulin. Biomolecules 2014, 4, 202–216. [CrossRef] 74. Arai, K.; Takei, T.; Shinozaki, R.; Noguchi, M.; Fujisawa, S.; Katayama, H.; Moroder, L.; Ando, S.; Okumura, M.; Inaba, K.; et al. Characterization and optimization of two-chain folding pathways of insulin via native chain assembly. Commun. Chem. 2018, 1, 1–11. [CrossRef] 75. Qiao, Z.-S.; Guo, Z.-Y.; Feng, Y.-M. Putative Disulfide-Forming Pathway of Porcine Insulin Precursor during Its Refolding in Vitro. Biochemistry 2001, 40, 2662–2668. [CrossRef][PubMed] 76. Qiao, Z.-S.; Min, C.-Y.; Hua, Q.-X.; Weiss, M.A.; Feng, Y.-M. In Vitro Refolding of Human Proinsulin. Kinetic intermediates, putative disulfide-forming pathway folding initiation site, and potential role of C-peptide in folding process. J. Biol. Chem. 2003, 278, 17800–17809. [CrossRef][PubMed] 77. Meienhofer, J.; Schnabel, E.; Bremer, H.; Brinkhoff, O.; Zabel, R.; Sroka, W.; Klostermayer, H.; Brandenburg, D.; Okuda, T.; Zahn, H. Synthesis of insulin and their combination to insulin Synthesis of insulin and their combination to insulin-active preparation. Z. Naturforsch. B 1963, 18, 1120–1121. [CrossRef] 78. Katsoyannis, P.G.; Fukuda, K.; Tometsko, A.; Suzuki, K.; Tilak, M. The Synthesis of the B-Chain of Insulin and Its Combination with Natural or Synthetis A-Chin to Generate Insulin Activity. J. Am. Chem. Soc. 1964, 86, 930–932. [CrossRef] 79. Kung, Y.; Du, Y.; Huang, W.; Chen, C.; Ke, L.; Hu, S.; Jiang, R.; Chu, S.; Niu, C.; Hsu, J.; et al. Total synthesis of crystalline insulin. Scientia Sinica 1966, 15, 544–561. [CrossRef] 80. Arai, K.; Takei, T.; Okumura, M.; Watanabe, S.; Amagai, Y.; Asahina, Y.; Moroder, L.; Hojo, H.; Inaba, K.; Iwaoka, M. Preparation of Selenoinsulin as a Long-Lasting Insulin Analogue. Angew. Chem. Int. Ed. 2017, 56, 5522–5526. [CrossRef] 81. Weil-Ktorza, O.; Rege, N.; Lansky, S.; Shalev, D.E.; Shoham, G.; Weiss, M.A.; Metanis, N. Substitution of an Internal Disulfide Bridge with a Diselenide Enhances both Foldability and Stability of Human Insulin. Chem. Eur. J. 2019, 25, 8513–8521. [CrossRef] 82. Polverino de Laureto, P.; Frare, E.; Gottardo, R.; van Dael, H.; Fontana, A. Partly folded states of members of the lysozyme/lactalbumin superfamily: A comparative study by circular dichroism spectroscopy and limited proteolysis. Protein Sci. 2002, 11, 2932–2946. [CrossRef] 83. Okumura, M.; Watanabe, S.; Inaba, K. CHAPTER 3.3: Structural Insights into Disulfide Bond Formation and Protein Quality Control in the Mammalian Endoplasmic Reticulum. In Oxidative Folding of Proteins; RSC Publishing: Cambridge, UK, 2018; pp. 224–248.