02.06.2013 Views

24 Peptides Synthesis of cystine-rich peptides - CHIMICA OGGI ...

24 Peptides Synthesis of cystine-rich peptides - CHIMICA OGGI ...

24 Peptides Synthesis of cystine-rich peptides - CHIMICA OGGI ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

<strong>Peptides</strong><br />

<strong>24</strong><br />

CYRIL BOULÈGUE<br />

HANS-JÜRGEN MUSIOL<br />

VIDYA PRASAD<br />

LUIS MORODER<br />

<strong>Synthesis</strong> <strong>of</strong> <strong>cystine</strong>-<strong>rich</strong> <strong>peptides</strong><br />

Abbreviations used: Acm, acetamidomethyl; Boc,<br />

tert-butoxycarbonyl; Bzl, benzyl; Bzl(4-Me),<br />

4-methylbenzyl; DPDS, di(2-pyridyl)disulfide; DMSO,<br />

dimethylsulfoxide; Fmoc, 9-fluorenylmethoxycarbonyl;<br />

Mob, 4-methoxybenzyl; Npys, 3-nitro-2-pyridylsulfanyl;<br />

Pys, 2-pyridylsulfanyl; StBu, tert-butylsulfanyl; sec,<br />

selenocysteine; tBu, tert-butyl; TFA, trifluoroacetic acid;<br />

TfOH, trifluoromethanesulfonic acid; Trt, trityl; Xan,<br />

9H-xanthen-9-yl.<br />

Since the early days <strong>of</strong> peptide chemistry the synthesis<br />

<strong>of</strong> cysteine-containing <strong>peptides</strong> has been one <strong>of</strong> the<br />

most challenging tasks, primarily because <strong>of</strong> the<br />

difficulties involved in the formation <strong>of</strong> multiple<br />

regioselective disulfide bonds. To overcome these<br />

challenges, new protection strategies and selective thiol<br />

chemistry continue to be developed. Great advances<br />

have been achieved over the years as demonstrated by<br />

the numerous highly efficient syntheses <strong>of</strong> mono- and<br />

multiple stranded Cys-<strong>rich</strong> <strong>peptides</strong>. The state <strong>of</strong> the art<br />

Table 1. The most common thiol protecting groups used in peptide synthesis.<br />

Stabilities and cleavage conditions are taken from similar compilations (1, 4).<br />

in the field has been extensively reviewed in recent<br />

years (1-5). The main strategies that have evolved for<br />

the synthesis <strong>of</strong> Cys-<strong>rich</strong> <strong>peptides</strong> can be classified as<br />

follows: i) stepwise regioselective Cys pairings; ii)<br />

convergent strategies based on a combination <strong>of</strong><br />

statistical oxidation <strong>of</strong> a minor number <strong>of</strong> Cys residues<br />

and regioselective disulfide formation; iii) oxidative<br />

folding <strong>of</strong> Cys-<strong>rich</strong> <strong>peptides</strong> by exploiting the structural<br />

information encoded in the sequence composition <strong>of</strong><br />

the target <strong>peptides</strong>; iv) induction <strong>of</strong> disulfide<br />

connectivities with selenocysteine (Sec). In this<br />

minireview we will summarize the achievements to<br />

date <strong>of</strong> the research in this developing field, highlight<br />

new insights that have emerged and address new<br />

perspectives. It is important to note that the results<br />

reviewed here are based on prior work and aims to<br />

summarize achievements <strong>of</strong> the recent years. Due to<br />

the breadth <strong>of</strong> this topic, this review can by no means<br />

be comprehensive. We intend to draw more general<br />

criteria and principles as future perspectives from the<br />

conventional wisdom acquired in the field.<br />

SYNTHESIS OF LINEAR CYSTEINE-<br />

PEPTIDE PRECURSORS<br />

Among the large number <strong>of</strong> thiol-protecting<br />

groups commonly used nowadays (Table 1) only<br />

a restricted set fulfils the requirements <strong>of</strong><br />

orthogonality in terms <strong>of</strong> the overall protection<br />

strategy (Fmoc/tBu or Boc/Bzl) and <strong>of</strong> reciprocal<br />

selectivity. The acid-sensitivity <strong>of</strong> many thiolprotecting<br />

groups and thus incompatibility with<br />

Boc/Bzl, has made Fmoc/tBu chemistry more<br />

applicable for larger combinatorial diversity. A<br />

major drawback is the repetitive piperidine<br />

treatments which are known to cause serious<br />

side reactions such as β-elimination and<br />

racemization. Racemization is particularly<br />

observed in the case when C-terminal Cys<br />

residues have to be directly coupled to the resin<br />

linker (3, 4, 6-12), e.g. relaxins and numerous<br />

toxins. In order to minimize these side reactions<br />

the thiol protecting groups Trt and Xan have<br />

proven to be more advantageous than Acm and<br />

StBu. An alternative strategy involves anchoring<br />

<strong>of</strong> the C-terminal Cys residues to the resin at the<br />

thiol function via the S-trityl (13) or the S-5-(9Hxanthen-9-yl-2-oxy)valeric<br />

acid anchor (14) as<br />

acid-labile linkers. This route provides effective<br />

access to free cysteine and <strong>cystine</strong> intermediates<br />

and/or products where both β-elimination and<br />

racemization are minimized. The former side<br />

reaction when occurring at the level <strong>of</strong> the sidechain<br />

anchored <strong>peptides</strong> leads to cleavage <strong>of</strong><br />

the growing peptide chain from the resin and<br />

consequently removal <strong>of</strong> the side product in the<br />

washing steps <strong>of</strong> the stepwise synthesis.<br />

chimica oggi • Chemistry Today • Vol <strong>24</strong> nr 4 • July/August 2006


REGIOSELECTIVE DISULFIDE FORMATION<br />

Intra- and intermolecular multiple disulfide bonds with the<br />

correct Cys connectivities are generated either by simple<br />

oxidative refolding after removal <strong>of</strong> a single-type thiol<br />

protecting group from synthetic precursors or by<br />

regioselective methods based on orthogonal thiolprotection<br />

schemes. The success <strong>of</strong> the oxidative refolding<br />

procedures strongly depends on the structural information<br />

retained by the Cys-<strong>rich</strong> <strong>peptides</strong> that in vivo are<br />

generally derived post-translationally from larger<br />

precursor forms, and thus on the thermodynamic stability<br />

<strong>of</strong> the target peptide. For the synthesis <strong>of</strong> such naturally<br />

occurring Cys-<strong>rich</strong> <strong>peptides</strong> and proteins, the oxidative<br />

refolding strategy is most <strong>of</strong>ten applied. However, even in<br />

such cases the correct <strong>cystine</strong>-network is not always<br />

formed. Therefore, for the synthesis <strong>of</strong> these natural<br />

products as well as for non-natural isomers or for de novo<br />

designed constructs an efficient chemical control in the<br />

generation <strong>of</strong> multiple intra- and intermolecular disulfide<br />

bonds is required. Due to the disappointingly low yields<br />

in the early efforts to obtain correctly folded insulin from<br />

the direct oxidative assembly <strong>of</strong> its A and B chains (15-<br />

18), great efforts have been spent in the following<br />

decades to develop orthogonal thiol protection<br />

schemes and related cleavage/oxidation chemistries<br />

needed for the stepwise regioselective formation <strong>of</strong><br />

disulfide bonds in single-, double- and even multiplestranded<br />

Cys-<strong>rich</strong> <strong>peptides</strong>. The advances in these<br />

rather complex multiple-step procedures have been<br />

comprehensively reviewed elsewhere (3, 4).<br />

A first milestone in this regioselective approach was<br />

achieved with the successful synthesis <strong>of</strong> the doublestranded<br />

human insulin where condensation <strong>of</strong><br />

fragments with preformed disulfide bridges was<br />

combined with regioselective cysteine-pairing<br />

procedures (19, 20). Subsequently the synthetic<br />

strategies have been further improved such that<br />

assembly <strong>of</strong> insulin and insulin-like <strong>peptides</strong> (21-23)<br />

and relaxins (<strong>24</strong>-26) from the two chains by<br />

regioselective intra- and interchain disulfide bridging<br />

methods can be carried out for both the Boc/Bzl and<br />

Fmoc/tBu synthetic strategies (Scheme 1).<br />

In the Boc/Bzl synthesis <strong>of</strong> the A and B chain <strong>of</strong><br />

insulin-4 (placentin) the transient thiol protecting<br />

group, i.e. Bzl(4-Me), is removed during the resincleavage<br />

and deprotection step using HF at 0-4 °C,<br />

i.e. conditions where the Cys(tBu) derivative would<br />

be expected to be stable. However, partial removal<br />

<strong>of</strong> the latter protecting group was observed<br />

confirming the non-ideal orthogonality between the<br />

Cys(tBu) and Cys(Bzl(4-Me)) derivatives. The initial<br />

intrachain disulfide bridging was achieved using<br />

DPDS in aqueous solution. This was followed by the<br />

removal <strong>of</strong> the tBu protecting group with in situ<br />

activation <strong>of</strong> the resulting free thiol by treatment with<br />

TfOH in TFA and in the presence <strong>of</strong> DPDS as proposed by<br />

Maruyama et al. (21). Subsequent interchain disulfide<br />

bridgings were formed by thiolysis under slightly basic<br />

conditions, followed by I 2-mediated<br />

deprotection/oxidation <strong>of</strong> the pair <strong>of</strong> Cys(Acm) residues.<br />

Although these basic conditions are not optimal for the<br />

thiolysis step due to the possibility <strong>of</strong> scrambling <strong>of</strong> the<br />

preformed intrachain disulfide bridge, side reactions <strong>of</strong><br />

this nature were not reported (23). Furthermore neither<br />

oxidation <strong>of</strong> the Met and Trp residues in the B chain <strong>of</strong><br />

insulin-4 (both these sensitive residues are absent in<br />

human insulin) nor cyclization at the indole group via<br />

thioether formation by intermediately formed Cys<br />

S-iodide were reportedly observed. Conversely,<br />

adaptation <strong>of</strong> the protection scheme reported by Akaji et<br />

al. for the synthesis <strong>of</strong> human insulin (22) to the synthesis<br />

<strong>of</strong> relaxin-3 by the Fmoc/tBu chemistry met with little<br />

success (26). Deprotection <strong>of</strong> the pair <strong>of</strong> Cys(tBu)<br />

derivatives and simultaneous disulfide bond formation by<br />

treatment with MeSiCl 3/PhS(O)Ph in TFA (27) caused<br />

substantial destruction <strong>of</strong> the Trp residues present in each<br />

chain <strong>of</strong> relaxin. This result is not surprising considering<br />

the near quantitative chlorination <strong>of</strong> indole groups under<br />

such conditions (28). Additionally, the position <strong>of</strong> the<br />

Cys(Acm) derivatives relative to the Trp residues in both<br />

chains was found to be critical in the iodine-mediated<br />

disulfide formation. Placing this protecting group in the<br />

proximity <strong>of</strong> the Trp in the A chain iodolysis led<br />

predominantly to modification <strong>of</strong> the indole side chain,<br />

whereas placing the two Acm groups at C-terminal<br />

positions, as shown in Scheme 1 (right panel), improved<br />

the final product yields (26). This observation fully<br />

confirms previous reports about the decisive role <strong>of</strong> the<br />

ring size that favors backbone-to-side chain cyclization<br />

via 2-thioether formation (29). Correspondingly for<br />

I 2-mediated Cys(Acm) cleavage and oxidation in the<br />

presence <strong>of</strong> Met and particularly Trp residues, careful<br />

Scheme 1. <strong>Synthesis</strong> <strong>of</strong> human insulin-4 (placentin) (left) by the Boc/Bzl (23)<br />

and relaxin-3 (right) by the Fmoc/tBu chemistry (26). The thiol protection<br />

strategies were adapted from the previous syntheses <strong>of</strong> the insulin-like insect<br />

peptide bombyxin IV (21) and human insulin (22). Regioselective disulfide<br />

formation was achieved by successive reactions with (i) DPDS in aqueous<br />

solution (pH 8.5), (ii) TfOH/TFA (1:5)/PhSMe (1 per cent)/DPDS at 0 °C,<br />

(iii) thiolysis at pH 8.5 at rt, and (iv) I 2/aqueous AcOH.<br />

optimization <strong>of</strong> reaction conditions is required (30,31).<br />

Accordingly, for thiolysis <strong>of</strong> the Cys(Pys) derivatives in the<br />

presence <strong>of</strong> preformed disulfides, slightly acidic<br />

conditions are preferred over neutral or basic to prevent<br />

or at least to largely suppress undesired thiol-disulfide<br />

exchange reactions.<br />

REGIOSELECTIVE FORMATION OF MULTIPLE<br />

INTRAMOLECULAR DISULFIDES<br />

The strategy <strong>of</strong> judiciously positioning pairs <strong>of</strong> Cys<br />

residues such as S-Trt, S-tBu and S-Acm derivatives along<br />

the peptide chain is currently the most widely applied<br />

procedure for the stepwise regioselective formation <strong>of</strong><br />

<strong>Peptides</strong><br />

chimica oggi • Chemistry Today • Vol <strong>24</strong> nr 4 • July/August 2006 25


<strong>Peptides</strong><br />

28<br />

three disulfide bonds in double-stranded and in singlestranded<br />

<strong>peptides</strong> (32). Thus, not only the peptide<br />

sequence itself, but also the distribution <strong>of</strong> the thiol<br />

protecting groups was found to strongly affect the success<br />

and the final yields <strong>of</strong> the desired disulfide isomers (33,<br />

34). Similar sequence dependencies were observed (34)<br />

when applying the DMSO/TFA-mediated<br />

deprotection/oxidation procedure (35,36) in a<br />

temperature-dependent mode for a one-pot regioselective<br />

disulfide formation between pairs <strong>of</strong> Cys(tBu) and<br />

Cys(Bzl(4-Me)) residues (37). This method was efficiently<br />

applied in the synthesis <strong>of</strong> bacterial enterotoxin ST 5-18<br />

with three disulfide bonds (38). In this one-pot procedure<br />

formation <strong>of</strong> the last disulfide using DMSO in TFA is<br />

carried out at<br />

higher<br />

temperatures<br />

(60 °C).<br />

Although such<br />

conditions may<br />

be acceptable<br />

in particular<br />

cases, there is<br />

a high<br />

propensity for<br />

sensitive<br />

residues such<br />

as Met and Trp<br />

to be oxidized,<br />

Table 2. Most common combinations <strong>of</strong><br />

Cys protections for multiple regioselective<br />

stepwise disulfide formation in singlestranded<br />

Cys-<strong>rich</strong> <strong>peptides</strong>.<br />

given that the<br />

respective<br />

byproducts<br />

were already<br />

observed at<br />

room temperature (35). Additionally, cleavage <strong>of</strong> acid<br />

sensitive peptide bonds such as Asp-Pro can occur (34)<br />

and succinimide formation at Asp-Xaa sequences is<br />

expected to occur under these conditions. An alternative<br />

one-pot regioselective formation <strong>of</strong> two disulfide bridges<br />

is possible due to the significantly different rates <strong>of</strong> I 2mediated<br />

oxidation <strong>of</strong> Cys(Trt) and Cys(Acm) derivatives<br />

in different solvent systems (39). By using CHCl 3, CH 2Cl 2<br />

and 2,2,2-trifluoroethanol as solvents the Cys(Trt)<br />

derivative is cleaved within seconds, whereas for the<br />

Cys(Acm) derivatives the reaction takes hours to complete.<br />

The difference in reactivities have been successfully<br />

applied to the stepwise regioselective formation <strong>of</strong> two<br />

disulfides (1, 40, 41).<br />

Table 2 contains the most common combinations <strong>of</strong><br />

pairwise Cys protections for stepwise regioselective<br />

disulfide formation in <strong>peptides</strong> prepared by the Boc/Bzl<br />

or Fmoc/tBu chemistry. In most cases for the first disulfide<br />

bond, thiol protections are selected which are removed in<br />

the resin cleavage/deprotection step. Alternatively, the<br />

Cys(StBu) protection is used which is cleaved reductively,<br />

preferentially in solution, to form the first disulfide bond,<br />

because under such conditions already preformed<br />

disulfides are also reduced. The procedure most<br />

commonly applied for the first disulfide formation is<br />

oxidation by either molecular oxygen in aqueous or<br />

aqueous/organic media under slightly basic conditions or<br />

alternatively by DMSO (36, 42). The method using<br />

DMSO can be applied over a wide pH range <strong>of</strong> 1 to 8,<br />

whereby faster oxidation rates are obtained with DMSO<br />

in acidic media. Since DMSO is known to disrupt<br />

aggregates, higher DMSO concentrations may<br />

additionally serve to solubilize <strong>peptides</strong>. Furthermore,<br />

modifications at sensitive amino acids like Met, Trp and<br />

Tyr are not observed by this oxidation method in aqueous<br />

solutions when operating at room temperature. More<br />

recently, polymer-bound oxidation agents such as the<br />

Ellman´s reagent have been proposed to facilitate the<br />

work-up procedures with the additional advantage <strong>of</strong><br />

allowing oxidation under slightly acidic conditions (43,<br />

44). In acidic media di(2-pyridyl)disulfide, pH 4 (45) or<br />

di(4-pyridyl)disulfide, pH 1 (44) are more efficient. The<br />

latter reagent could possibly serve for intramolecular<br />

disulfide formation in the presence <strong>of</strong> an already<br />

preformed disulfide bond because thiol-disulfide<br />

exchange reactions are largely suppressed under acidic<br />

conditions. Therefore, oxidation <strong>of</strong> free thiols by di(4pyridyl)disulfide<br />

at pH values in the range <strong>of</strong> 1 to 4 could<br />

be advantageous over the DMSO/TFA procedure<br />

particularly in the presence <strong>of</strong> Met and Trp residues,<br />

thereby increasing the diversity <strong>of</strong> possible combinations<br />

<strong>of</strong> thiol-protecting groups.<br />

To our knowledge, there has been a limited number <strong>of</strong><br />

reports <strong>of</strong> stepwise regioselective formation <strong>of</strong> more than<br />

three disulfide bonds. This is most likely due to the<br />

observation that with increasing number <strong>of</strong> disulfide<br />

bridges, oxidative refolding is found to be sufficiently<br />

successful as final synthetic step. Nevertheless, at least from<br />

an academic point <strong>of</strong> view, stepwise regioselective<br />

formation <strong>of</strong> a larger number <strong>of</strong> disulfides can well be<br />

envisaged from the knowledge accumulated on orthogonal<br />

combinations <strong>of</strong> thiol protecting groups, as shown in<br />

Scheme 2. Indeed, a similar protection scheme without the<br />

Cys(StBu) pair was used for the stepwise formation <strong>of</strong> four<br />

disulfides in an α-conotoxin dimer by exploiting the<br />

different rates <strong>of</strong> Cys(tBu) and Cys(Bzl(4-Me)) oxidation<br />

with DMSO/TFA at different temperatures (46).<br />

Extension <strong>of</strong> the number <strong>of</strong> disulfides to five would involve<br />

the critical temperature-dependent cleavage and<br />

oxidation <strong>of</strong> the Cys(Bzl(4-Me)) derivatives with<br />

DMSO/TFA discussed previously. But such an extensively<br />

disulfide-crosslinked peptide is unlikely to be a synthetic<br />

target.<br />

Scheme 2. Potential thiol protection scheme for the<br />

regioselective formation <strong>of</strong> up to five intramolecular<br />

disulfide bonds: (i) reductive cleavage <strong>of</strong> the<br />

S-tert-butylsulfanyl groups and oxidation with<br />

di(2-pyridyl)disulfide or DMSO on resin or in<br />

aqueous/organic media upon cleavage <strong>of</strong> the fully<br />

protected peptide from the resin by mild acid<br />

conditions; (ii) deprotection <strong>of</strong> the peptide with TFA<br />

with concomitant removal <strong>of</strong> the S-trityl groups<br />

followed by HPLC purification and oxidation with<br />

DPDS under acidic conditions; (iii) iodine-mediated<br />

oxidation preferentially in aqueous AcOH; (iv)<br />

DMSO/TFA at rt, and (v) DMSO/TFA at 60 °C.<br />

chimica oggi • Chemistry Today • Vol <strong>24</strong> nr 4 • July/August 2006


Scheme 3. Disulfide crosslinking <strong>of</strong> collagenous <strong>peptides</strong> into<br />

heterotrimers by regioselective thiol chemistry.<br />

REGIOSELECTIVE FORMATION OF MULTIPLE<br />

INTERCHAIN DISULFIDES<br />

For homodimerization <strong>of</strong> a mono-cysteine peptide, the<br />

methodologies used for intramolecular cyclization <strong>of</strong><br />

<strong>peptides</strong> by a disulfide bridge can be employed. In the<br />

case <strong>of</strong> hetero-dimerization and particularly hetero-<br />

oligomerization, selective thiol chemistries are required.<br />

This generally involves thiolysis <strong>of</strong> an unsymmetric<br />

disulfide such as the Cys(Pys), Cys(Npys) or the (5-nitro-<br />

2-pyridyl)sulfanyl derivative with an unprotected Cys thiol<br />

group <strong>of</strong> the second strand (Scheme 3). The unsymmetric<br />

disulfide species are obtained by reaction <strong>of</strong> an<br />

unprotected Cys residue <strong>of</strong> one peptide strand with di(2pyridyl)disulfide<br />

and di(3-nitro-2-pyridyl)disulfide or by<br />

the related sulfanylchlorides (Pys-Cl or Npys-Cl) (4). The<br />

sulfanylchlorides serve also to convert a Cys(Acm) residue<br />

<strong>of</strong> one <strong>of</strong> the disulfide crosslinked strands into a Cys(Pys)<br />

or Cys(Npys) derivative for a further oligomerization step<br />

by thiolysis as shown in Scheme 3, whereby these<br />

reactions have to be performed under acidic conditions to<br />

prevent thiol-disulfide exchanges at preformed disulfide<br />

bridges (47). Since the sulfanylchlorides react at high<br />

rates even with indole groups to form the related 2thioethers,<br />

such a cascade <strong>of</strong> reactions for regioselective<br />

disulfide crossbridging <strong>of</strong> peptide ladders cannot be<br />

applied in the presence <strong>of</strong> Trp residues. A possible<br />

alternative could be the direct conversion <strong>of</strong> Cys(Acm)<br />

residues into Cys(Pys) or Cys(Npys) derivatives by<br />

reaction with iodine in the presence <strong>of</strong> di(2pyridyl)disulfide<br />

or di(3-nitro-2-pyridyl)disulfide, which<br />

immediately trap the intermediately formed sulfanyl<br />

iodide (48). Disulfide crosslinking <strong>of</strong> collagen <strong>peptides</strong><br />

into heterotrimers by artificial <strong>cystine</strong>-knots has been very<br />

successful (47, 49, 50), although yields were strongly<br />

dictated by conformational effects (51, 52).<br />

Conformational effects cannot be excluded for the failure<br />

<strong>of</strong> general strategies when applied to the production <strong>of</strong><br />

<strong>Peptides</strong><br />

chimica oggi • Chemistry Today • Vol <strong>24</strong> nr 4 • July/August 2006 29


<strong>Peptides</strong><br />

30<br />

Scheme 4. Semiselective formation <strong>of</strong> three disulfides by a<br />

two-step procedure based on statistical oxidation <strong>of</strong> four thiols<br />

by air oxygen or DMSO in aqueous solution, followed by<br />

regioselective I 2-mediated oxidation <strong>of</strong> the two purposely<br />

placed Cys(Acm) derivatives.<br />

multiple disulfide bridges in Cys-<strong>rich</strong> <strong>peptides</strong>. These<br />

effects may not occur when disulfide formation is<br />

performed under non-aqueous conditions, but can be<br />

more pronounced when carrying out the selective Cys<br />

pairing reactions in aqueous solutions. In fact, even<br />

shorter peptide sequences may retain sufficient structural<br />

information for local conformational preferences in water.<br />

On the other hand, these locally folded <strong>peptides</strong> can<br />

advantageously be exploited for combining oxidative<br />

folding procedures and regioselective disulfide formation<br />

(vide infra).<br />

SEMISELECTIVE DISULFIDE FORMATION<br />

Despite the orthogonality <strong>of</strong> the thiol-protecting groups<br />

and the methods developed for their oxidative<br />

deprotection allow for regioselective formation <strong>of</strong> three<br />

and even more disulfide bridges, such syntheses require<br />

careful planning because <strong>of</strong> the multiple sequential<br />

deprotection/oxidation and purification steps. In<br />

principle, formation <strong>of</strong> three disulfide bonds by a<br />

combination <strong>of</strong> two different oxidative procedures, e.g.<br />

air oxidation <strong>of</strong> four Cys residues followed by one<br />

regioselective iodine-mediated disulfide formation can<br />

also be envisaged. However, in practice separation <strong>of</strong><br />

Scheme 5. Two step oxidative refolding <strong>of</strong> synthetic hirudin 1-65 .<br />

isomers is inevitable because <strong>of</strong> the incomplete<br />

regioselectivity in the two disulfide bridging steps, as<br />

shown in Scheme 4 (32, 53). The advantage <strong>of</strong> this<br />

strategy is the relatively fast access to three well defined<br />

disulfide isomers instead <strong>of</strong> the 15 expected from random<br />

oxidation <strong>of</strong> six Cys residues. With the knowledge<br />

derived from families with common Cys-sequence patterns<br />

and hypothetically common disulfide connectivities, the<br />

two Cys(Acm) derivatives can be placed at defined<br />

positions to induce the desired isomers. These products<br />

can then be compared chromatographically with the<br />

natural compound. In the event that this material is not<br />

available, their bioactivities can be used to confirm or<br />

identify the native <strong>cystine</strong> frameworks. The procedure may<br />

also serve to resolve specific biological properties as well<br />

documented, e.g., in the case <strong>of</strong> β-defensins (54).<br />

Given that identical Cys-sequence patterns generally<br />

lead to identical <strong>cystine</strong> frameworks even in the case <strong>of</strong><br />

low sequence homology (vide infra), placing the<br />

unprotected Cys residues at positions for successive<br />

disulfide crosslinking <strong>of</strong> the most proximal thiols in the<br />

sequence and the two Cys(Acm) residues at positions for<br />

formation <strong>of</strong> a disulfide that crosses one <strong>of</strong> the<br />

preformed consecutive disulfide bridges (Scheme 5), the<br />

probability <strong>of</strong> preferential generation <strong>of</strong> a single isomer<br />

is significantly enhanced by the quasi-stochastic<br />

oxidative folding <strong>of</strong> Cys-<strong>rich</strong> <strong>peptides</strong> according to the<br />

“proximity rule” (55). With the power-law dependence<br />

on loop length during the fast early oxidation and<br />

reshuffling steps, the thiols encounter each other with a<br />

statistical probability determined primarily by the loop<br />

entropy, albeit possibly modified by local<br />

conformational biases in the unfolded state. Indeed,<br />

applying such knowledge-based design to the protection<br />

scheme, straightforward generation <strong>of</strong> the desired native<br />

framework has been achieved (33, 56-58). An example<br />

<strong>of</strong> the application potential <strong>of</strong> this rather simple synthetic<br />

strategy has been recently reported for the synthesis <strong>of</strong><br />

hirudin (56). Detailed analysis <strong>of</strong> the oxidative folding<br />

pathways <strong>of</strong> hirudin had clearly revealed a sequential<br />

flow <strong>of</strong> the unfolded chain through an ensemble <strong>of</strong><br />

equilibrated one-disulfide to ensembles <strong>of</strong> equilibrated<br />

two-disulfide intermediates, with no accumulation <strong>of</strong><br />

preferred native or non-native isomers. From these<br />

ensembles the correct native isomer accumulates either<br />

directly by formation <strong>of</strong> the third disulfide bridge or by<br />

reshuffling under reducing conditions. In this way the<br />

thermodynamically most favoured native <strong>cystine</strong><br />

framework is obtained in high yields (59-61).<br />

Accordingly, the exclusive formation <strong>of</strong> the correct bisdisulfide<br />

isomer with the sequential Cys 6 -Cys 14 , Cys 22 -<br />

Cys 39 disulfides by oxidation <strong>of</strong> the reduced synthetic<br />

Cys(Acm) 16 ,Cys(Acm) 28 -hirudin 1-65 in the presence <strong>of</strong> βmercaptoethanol<br />

(Scheme 5) was particularly surprising,<br />

albeit in agreement with the “proximity” rule. This<br />

intermediate was then converted in high yields by I 2oxidation<br />

<strong>of</strong> the Cys(Acm) pair into the native <strong>cystine</strong><br />

framework <strong>of</strong> hirudin by formation <strong>of</strong> the crossing<br />

disulfide bridge. This strategy in other cases however,<br />

proved to be rather unsuccessful as e.g. in ref. (34).<br />

OXIDATIVE FOLDING<br />

Anfinsen proposed the concept that an intrinsic folding<br />

code determines the correct folding <strong>of</strong> proteins into the<br />

presumed thermodynamic ground states uniquely<br />

governed by the amino acid sequence (62). This basic<br />

concept also applies when folding is coupled to disulfide<br />

formation, because the native conformation is only<br />

chimica oggi • Chemistry Today • Vol <strong>24</strong> nr 4 • July/August 2006


stabilized, not specified by the disulfide bonds. The<br />

oxidation <strong>of</strong> Cys residues to disulfides is a priori<br />

undirected, rather it requires spatial proximity <strong>of</strong> these<br />

residues. While conformational preferences can promote<br />

contacts and, thus, pairing <strong>of</strong> Cys residues, the formation<br />

<strong>of</strong> a disulfide bond strongly stabilizes the threedimensional<br />

structure. Folding in vivo is supported by a<br />

wide variety <strong>of</strong> molecular chaperons and <strong>of</strong> folding<br />

catalysts. These do not determine the final conformation<br />

<strong>of</strong> the polypeptide chain, but rather increase the efficiency<br />

<strong>of</strong> the folding process by inhibiting <strong>of</strong>f-pathway<br />

aggregation phenomena, catalyzing reshuffling <strong>of</strong><br />

disulfides and rate-limiting isomerization steps. The<br />

critical interplay between the coupled processes <strong>of</strong> folding<br />

and oxidation becomes most evident in Cys-<strong>rich</strong> bioactive<br />

<strong>peptides</strong> and miniproteins. In fact, synthetic replicates<br />

<strong>of</strong>ten fold preferentially into the native disulfide isomers<br />

under optimized oxidative conditions, even though the<br />

reduced forms are mostly unstructured (63).<br />

An ever growing number <strong>of</strong> Cys-<strong>rich</strong> <strong>peptides</strong> were<br />

discovered and isolated in recent years from the most<br />

diverse kingdoms <strong>of</strong> life including mammalians. These<br />

<strong>peptides</strong> include hormones, growth factors, protease<br />

inhibitors, components <strong>of</strong> the innate immunity system,<br />

toxins, antimicrobial agents etc. In the biosynthesis, these<br />

Cys-<strong>rich</strong> <strong>peptides</strong> are produced ribosomally as precursor<br />

molecules containing N- and C-terminal prosequences<br />

that can act as intramolecular chaperones during the<br />

folding process and are cleaved in the post-translational<br />

maturation process from the prefolded precursors to<br />

produce single- and in selected cases double-stranded<br />

disulfide-<strong>rich</strong> <strong>peptides</strong> <strong>of</strong> varying lengths from about 10<br />

to 60-70 amino acid residues. Moreover, in addition to<br />

the conformationally restricting disulfide bonds, backbone<br />

cyclizations have also been observed in plant protease<br />

inhibitors or plant and mammalian antimicrobial <strong>peptides</strong><br />

(64-66).<br />

The production <strong>of</strong> these bioactive Cys-<strong>rich</strong> <strong>peptides</strong> by<br />

recombinant technologies using different host organisms<br />

is a challenging task, because <strong>of</strong> major obstacles such as<br />

low expression rates, high susceptibility towards<br />

degradation by the host cell proteases and a significant<br />

toxicity for the host organisms. Hence, chemical synthesis<br />

still represents the main access route. However, as<br />

products <strong>of</strong> post-translational maturation processes these<br />

<strong>peptides</strong> may have partially or totally lost the essential<br />

structural information for correct oxidative refolding.<br />

Consequently, a drastically decreased efficiency is likely<br />

for such folding processes given the number <strong>of</strong> possible<br />

isomers formed by random oxidation with increasing Cys<br />

residues. Indeed with four, six or eight Cys residues in<br />

principle 3, 15, and 105 different intramolecular disulfide<br />

isomers can be generated. Furthermore, the complexity <strong>of</strong><br />

possible isomers increases enormously when two or more<br />

polypeptide chains are crosslinked by interchain disulfide<br />

bridges as in the classical case <strong>of</strong> the insulin family <strong>of</strong><br />

hormones (67). For these double-stranded poly<strong>peptides</strong><br />

the theoretically expected yield from random oxidation <strong>of</strong><br />

the two chains should therefore be close to zero;<br />

nevertheless under optimized conditions human insulin is<br />

obtained in yields <strong>of</strong> up to 25-30 percent (68, 69). This<br />

result suggests that the native fold represents the most<br />

stable among the possible isomers (67). Equally successful<br />

was the assembly <strong>of</strong> relaxin-2 from the A and B chain<br />

(70, 71). On the contrary, this oxidative refolding<br />

procedure failed completely in the case <strong>of</strong> relaxin-3 (26)<br />

or insulin-4 (23), suggesting fine-tuned conformational<br />

preferences to be responsible for success or failure.<br />

An increased number <strong>of</strong> disulfide bridges in relatively<br />

short polypeptide chains leads to compaction <strong>of</strong> the<br />

globular structures with the disulfides mainly buried in the<br />

nonpolar core. This implies that such mature protein<br />

fragments should represent even in the precursor<br />

molecules stable subdomains. Therefore, at least to some<br />

extent sufficient structural information may be retained in<br />

their sequence for correct folding, provided appropriate<br />

experimental conditions are applied in terms <strong>of</strong> peptide<br />

concentration to prevent oligomerization, redox reagents,<br />

pH values, temperature, additives for solubilization or<br />

detergents as well as aqueous/organic media for both<br />

solubilization and conformational stabilization (63).<br />

Due to the varying yields <strong>of</strong> oxidative refolding among<br />

families and even within families <strong>of</strong> these Cys-<strong>rich</strong><br />

<strong>peptides</strong>, an important role was speculatively assigned to<br />

the propeptide sequence (72). Indeed, propeptidefacilitated<br />

oxidative folding has been observed for a<br />

number <strong>of</strong> Cys-<strong>rich</strong> molecules, but there is still debate as<br />

to whether the prosequences or more importantly<br />

particular intramolecular and/or intermolecular<br />

interactions with specialized structural elements promote<br />

appropriate conformations to extents that support the<br />

correct oxidative refolding into the native disulfide<br />

frameworks (72-82). The question about the role <strong>of</strong> the<br />

prosequences is particularly pertinent for <strong>peptides</strong> with<br />

identical Cys-sequence patterns that fold into identical<br />

disulfide frameworks despite their marked sequence<br />

variability (83-85). In the case <strong>of</strong> the “conotoxin folding<br />

puzzle”(83) the propeptide portion plays a role in the<br />

PDI-assisted folding. In addition, particular posttranslational<br />

modifications such as γ-carboxylation <strong>of</strong><br />

glutamate residues or C-terminal amidation were found to<br />

affect yields <strong>of</strong> oxidative folding (84, 86). The<br />

experimental benefit observed with pro<strong>peptides</strong> in vitro<br />

may well rely on the intramolecular chaperone-like<br />

activity <strong>of</strong> the prosequences in terms <strong>of</strong> increased<br />

solubility <strong>of</strong> folding precursors which suppresses<br />

aggregation phenomena and thus precipitation. Upon<br />

oxidative folding, generally an enhanced solubility is<br />

observed by burial <strong>of</strong> hydrophobic patches into the core<br />

<strong>of</strong> the usually compact globular folds <strong>of</strong> the Cys-<strong>rich</strong><br />

micro- and miniproteins.<br />

Although the production <strong>of</strong> native disulfide isomers in<br />

acceptable yields <strong>of</strong>ten fails, surprisingly an equal<br />

number <strong>of</strong> high rates <strong>of</strong> success are reported. However,<br />

the intrinsic driving forces encoded in these short<br />

peptide sequences for the unidirectional folding into<br />

distinct disulfide frameworks still remain a biochemical<br />

and structural puzzle. Stabilization <strong>of</strong> preferred ordered<br />

structures such as α-helices or β-sheets in structural<br />

motifs by distinct disulfide bonds may well represent the<br />

Figure 1. Possible disulfide isomers <strong>of</strong> <strong>peptides</strong> containing four Cys residues<br />

(top) and the preferred isomers formed upon oxidative folding <strong>of</strong> apamin<br />

and endothelin-1 (bottom).<br />

<strong>Peptides</strong><br />

chimica oggi • Chemistry Today • Vol <strong>24</strong> nr 4 • July/August 2006 31


32<br />

Figure 2. NMR structures<br />

<strong>of</strong> apamin and related<br />

seleno<strong>cystine</strong> analogues:<br />

A) wild-type Cys 1-11 ,Cys 3-15 -<br />

apamin;<br />

B) Sec 1-11 ,Cys 3-15 -apamin with<br />

the wild-type connectivities<br />

(globule isomer);<br />

C) Cys 1-15 ,Sec 3-11 -apamin<br />

(ribbon isomer) as main<br />

conformer with trans Pro 6 ;<br />

D) Sec 1-11 ,Cys 3-15 -apamin<br />

(ribbon isomer) with cis Pro 6<br />

(20 percent); E) Sec 1-3 ,Cys 11-15 -<br />

apamin isomer (bead isomer).<br />

Figure 3: Primary structure <strong>of</strong> minicollagen-1 from Hydra consisting <strong>of</strong> a<br />

propeptide [1-9], the N-terminal Cys-<strong>rich</strong> domain [10-32], poly-Pro<br />

[33-55], Gly-Pro-Pro triplets [60-101], poly-Pro [103-108] and the<br />

C-terminal Cys-<strong>rich</strong> domain [108-130].<br />

major driving force. Indeed, a clear correlation between Cys-sequence<br />

patterns, disulfide networks and consequently, overall folds has been<br />

observed despite the different origins and functions <strong>of</strong> these Cys-<strong>rich</strong><br />

<strong>peptides</strong>. This observation is made even in cases <strong>of</strong> poor sequence<br />

homology in the non-cysteine residues (87-94). Among these recurrent<br />

folds the most common structural motifs are the <strong>cystine</strong>-stabilized αβ, the<br />

<strong>cystine</strong> knot (knottin) and the β-hairpin-like motifs. Therefore, from the large<br />

body <strong>of</strong> data collected with the Cys-<strong>rich</strong> <strong>peptides</strong> the general consensus<br />

evolved, that identical Cys-sequence patterns correlate with identical<br />

disulfide connectivities and thus identical structural folds. Consequently, at<br />

least within superfamilies <strong>of</strong> Cys-<strong>rich</strong> <strong>peptides</strong> with identical Cys-sequence<br />

patterns it is assumed that the disulfide connectivities are identical and the<br />

overall folds very similar despite the low degree <strong>of</strong> sequence homology.<br />

Unfortunately however, only few investigations were carried out to prove<br />

experimentally the validity <strong>of</strong> these assumptions by either determining<br />

directly the disulfide connectivities <strong>of</strong> the natural products or by comparison<br />

with synthetic replicates. Therefore, assignment <strong>of</strong> disulfide connectivities by<br />

simple analogy should still be handled with caution as clearly illustrated by<br />

the two examples discussed below.<br />

The Cys-Xaa-Cys/Cys-(Xaa) 3-Cys sequence pattern<br />

Independent <strong>of</strong> the disulfide connectivities which may be dictated by<br />

Figure 4. Disulfide connectivities and solution structures <strong>of</strong> the N- and<br />

C-terminal Cys-<strong>rich</strong> domain <strong>of</strong> minicollagen-1 from Hydra nematocysts.<br />

chimica oggi • Chemistry Today • Vol <strong>24</strong> nr 4 • July/August 2006


<strong>Peptides</strong><br />

34<br />

additional Cys residues present in the sequence,<br />

almost all the <strong>peptides</strong> and proteins that contain the<br />

Cys-(Xaa) 1-Cys/Cys-(Xaa) 3-Cys pattern, show the<br />

characteristic <strong>cystine</strong>-stabilized αβ structural motif,<br />

except for the <strong>cystine</strong>-knot <strong>of</strong> the growth factor<br />

superfamily that exhibits a four-stranded irregular<br />

antiparallel β sheet where the ring formed by the two<br />

disulfide bridges is penetrated by a third disulfide (87).<br />

The bee venom toxin apamin and the human hormone<br />

endothelin-1 exhibit the identical Cys pattern Cys-<br />

(Xaa) 1-Cys-(Xaa) 7-Cys-(Xaa) 3-Cys (Figure 1).<br />

Considering the possible disulfide connectivities, the<br />

two peptide backbones can be crosslinked by two<br />

disulfides in a parallel (ribbon isomer), crossing<br />

(globule isomer) or in a sequential manner (beads<br />

isomer). Oxidative refolding <strong>of</strong> the apamin generates<br />

quantitatively the native globule isomer (95), while in<br />

the case <strong>of</strong> endothelin-1 a mixture is formed <strong>of</strong> the<br />

native ribbon and non-native globule isomer at a ratio<br />

<strong>of</strong> 3:1 (96). Despite the identical Cys pattern, two<br />

different disulfide frameworks are preferred which<br />

nonetheless lead to the identical αβ fold (87). In both<br />

cases a sequential disulfide pattern (bead isomer) has<br />

not been observed in the oxidative folding mixtures, a<br />

fact that fully agrees with the disfavoured ring sizes <strong>of</strong><br />

disulfide-bridged loops with one or three residues<br />

spacing the cysteines (97).<br />

In order to disclose the structural factors responsible<br />

for the distinct oxidative folding behaviour <strong>of</strong><br />

apamin and endothelin, the highly reductive redox<br />

potential <strong>of</strong> Sec (98) was exploited to replace<br />

isosterically at three defined positions one disulfide<br />

with a diselenide group. Using this procedure, all<br />

three possible apamin isomers were obtained by<br />

oxidation <strong>of</strong> the two residual Cys residues (Figure 2)<br />

(99, 100). The preferred 3D structure <strong>of</strong> the Sec 1-<br />

11 ,Cys 3-15 -apamin analogue as globule isomer fully<br />

confirms the isosteric character <strong>of</strong> a diselenide. This<br />

makes such disulfide replacements particularly suited<br />

for the synthesis <strong>of</strong> heavy metal analogues <strong>of</strong><br />

<strong>peptides</strong> and proteins (101). In addition, it could<br />

serve to enhance the robustness <strong>of</strong> disulfide<br />

frameworks especially because the reduction <strong>of</strong><br />

diselenides by thiols <strong>of</strong> redox potentials in the range<br />

<strong>of</strong> glutathione is difficult to achieve (102). In the<br />

ribbon isomer the main conformer retains almost the<br />

identical overall fold as the wild-type apamin isomer<br />

with the Pro residue <strong>of</strong> the trans Ala 5 -Pro 6 buried<br />

together with the hydrophobic diselenide in the core<br />

<strong>of</strong> the molecule. A second conformer present in<br />

solution (20 per cent), however, exhibits an<br />

endothelin-like structure which is apparently induced<br />

by the cis Ala 5 -Pro 6 conformation that exposes the<br />

Pro side chain to the bulk <strong>of</strong> the solvent. Recently, an<br />

interesting finding was reported where the<br />

preference <strong>of</strong> cis or trans Xaa-Pro bonds strongly<br />

depends on the size <strong>of</strong> disulfide loops (103). The<br />

readiness to form a disulfide loop is known to<br />

depend on the number <strong>of</strong> intervening residues m<br />

between the two cysteines which is more important<br />

than the type <strong>of</strong> residues at least up to m = 6, with<br />

odd values <strong>of</strong> m being disfavoured and even<br />

numbers favouring the ring closure (97). In terms <strong>of</strong><br />

cis and trans aminoacyl-Pro conformations the<br />

opposite rule was observed, with odd m numbers<br />

favouring in distinct manner the cis conformation<br />

(103). This may well account for the presence <strong>of</strong> the<br />

second conformer <strong>of</strong> Cys 1-15 ,Sec 3-11 apamin with m<br />

= 7 for the Sec 3-11 loop, but the resulting less<br />

compact structure should be thermodynamically less<br />

favoured than the wild-type globule isomer and thus<br />

less populated. Finally, the structure <strong>of</strong> the bead<br />

isomer clearly confirmed that a transient formation<br />

<strong>of</strong> this isomer in the early fast steps <strong>of</strong> oxidative<br />

folding according to the proximity rule would induce<br />

formation <strong>of</strong> a distorted α-helix which relaxes into<br />

the α-helix <strong>of</strong> the wild-type isomer upon thioldisulfide<br />

exchange reactions involving the<br />

disfavoured N-terminal Cys-Xaa-Cys loop. Thus, the<br />

proline residue in apamin, which is absent in<br />

endothelin-1, may well be the principal cause <strong>of</strong><br />

their different disulfide frameworks, providing an<br />

exception to the general rules discussed above.<br />

Cysteine-<strong>rich</strong> domains <strong>of</strong> minicollagen-1<br />

from hydra nematocysts<br />

A similar decisive role <strong>of</strong> a proline residue in<br />

dictating the disulfide connectivities in <strong>peptides</strong> with<br />

identical Cys-sequence pattern was recently observed<br />

for the Cys-<strong>rich</strong> subdomains <strong>of</strong> minicollagen-1 from<br />

Hydra nematocysts. As shown in Figure 3, the Nand<br />

C-terminal domains <strong>of</strong> this collagen, which most<br />

probably derive from gene duplication, are<br />

characterized by the identical Cys pattern and by<br />

one conserved proline, while all other constituent<br />

residues differ in the two domains (104).<br />

Although the oxidative folding rates <strong>of</strong> synthetic<br />

replicates <strong>of</strong> the two domains in the presence <strong>of</strong><br />

oxidized/reduced glutathione (9:1) and at pH 8.0<br />

differ significantly, in both cases almost exclusively one<br />

disulfide isomer was obtained. The disulfide<br />

connectivities <strong>of</strong> these two main oxidation products<br />

were derived unambiguously from their NMR<br />

structures in solution (105,106). Despite the identical<br />

Cys-sequence pattern, two different disulfide<br />

frameworks were determined and correspondingly two<br />

differing solution structures, as shown in Figure 4.<br />

Most surprising were the different conformations <strong>of</strong> the<br />

Xaa-Pro bond involving the conserved proline residue<br />

in the two domains. Detailed studies <strong>of</strong> the folding<br />

pathways clearly revealed an unidirectional folding<br />

pathway from the reduced to the correctly folded<br />

isomer with the trans Xaa-Pro 119 bond for the Cterminal<br />

domain without the transient formation <strong>of</strong><br />

intermediates (107). Conversely, for the N-terminal<br />

domain accumulation <strong>of</strong> a fully oxidized intermediate<br />

was detected with all six Cys residues disulfide-linked<br />

in a sequential manner, i.e. with the two disfavoured<br />

Cys-(Xaa) 3-Cys loops and the even more disfavoured<br />

C-terminal vicinal Cys-Cys loop. The disulfide<br />

connectivities <strong>of</strong> this intermediate, the conversion <strong>of</strong> the<br />

trans to cis Ala-Pro 25 bond in the transiently formed<br />

Cys-Ala-Pro-Val-Cys ring structure (in agreement with<br />

ref. (103)) and additional information derived from<br />

related (4R)- and (4S)-FPro analogues allowed to<br />

unequivocally assign the decisive role in dictating the<br />

folding pathway to the conserved proline.<br />

PERSPECTIVES<br />

From the collective experience gained over the years<br />

in the synthesis <strong>of</strong> Cys-<strong>rich</strong> <strong>peptides</strong> by using in the<br />

last step either regioselective thiol chemistries or the<br />

sequence-encoded structural information to establish<br />

the desired disulfide connectivities, it can be<br />

concluded that the main drawback stems from the<br />

individual behaviour <strong>of</strong> each peptide which limits<br />

general procedures to be proposed and applied.<br />

Nevertheless, the oxidative folding approach remains<br />

chimica oggi • Chemistry Today • Vol <strong>24</strong> nr 4 • July/August 2006


attractive because <strong>of</strong> its simplicity compared to the<br />

synthetic strategies for regioselective disulfide bond<br />

formation. It is certainly indispensable if the number<br />

<strong>of</strong> Cys residues exceeds the presently available<br />

chemistry for site-directed Cys pairings. With careful<br />

attention paid to experimental conditions, the nativetype<br />

fold <strong>of</strong> the peptide can be accomplished;<br />

however, misfolded disulfide isomers are <strong>of</strong>ten<br />

produced as main products in spite <strong>of</strong> continuous<br />

efforts to optimize the reaction conditions. Therefore,<br />

great attention has to be paid to experimental<br />

conditions such as pH, buffer, peptide concentration,<br />

redox reagents, temperature, additives, denaturants<br />

and organic solvents to generate the desired disulfide<br />

connectivities in more satisfying yields (63). In this<br />

context, besides the traditional use <strong>of</strong> glutathione as<br />

redox agent, increasingly the <strong>cystine</strong>/cysteine pair is<br />

employed and other thiols have been investigated<br />

such as aromatic thiols (108,109) or dithiols (110).<br />

Moreover, additives like guanidinium hydrochloride<br />

at low concentrations or detergents (111) proved to<br />

be useful for preventing aggregation and thus,<br />

oligomerization. However, based on the assumption<br />

that within families or even superfamilies <strong>of</strong> Cys-<strong>rich</strong><br />

<strong>peptides</strong> the predominant isomer formed by the<br />

oxidative folding procedures should represent the<br />

native isomer, the experimental pro<strong>of</strong> <strong>of</strong> the Cys<br />

connectivities has been largely neglected, although<br />

with the fast advancements in mass-spectrometry<br />

such analytical problems can be readily resolved.<br />

Since nature ingeniously applies complex disulfide<br />

frameworks to stabilize the most diverse spatial<br />

arrangements <strong>of</strong> epitopes for efficient cross-talking<br />

with receptor molecules, it is highly tempting to mimic<br />

such natural scaffolds for the engineering <strong>of</strong><br />

miniature proteins in which functional sequence<br />

portions <strong>of</strong> other bioactive <strong>peptides</strong> are suitably<br />

displayed for molecular recognition processes.<br />

Indeed, the structural robustness <strong>of</strong> the disulfide<br />

frameworks <strong>of</strong> natural Cys-<strong>rich</strong> families <strong>of</strong> peptide<br />

which is well demonstrated by the large diversity <strong>of</strong><br />

the sequence compositions in the non-cysteine<br />

positions, has already been exploited with success for<br />

the generation <strong>of</strong> hybrid molecules (112, 113). Such<br />

new type <strong>of</strong> de novo design <strong>of</strong> bioactive molecules by<br />

the use <strong>of</strong> natural disulfide frameworks may possibly<br />

further advance with the selenocysteine approach.<br />

This promising strategy is based on replacement <strong>of</strong><br />

Cys pairs involved in the formation <strong>of</strong> unproductive<br />

folding intermediates by Sec residues which force<br />

diselenide formation at critical disulfide positions,<br />

resulting in the suppression <strong>of</strong> unproductive<br />

kinetically trapped intermediates. By such an<br />

approach not only significantly improved yields <strong>of</strong><br />

the desired disulfide/diselenide frameworks can be<br />

expected, but also the access to biopharmaceuticals<br />

<strong>of</strong> superior stability under biological reducing<br />

conditions, if all disulfides are replaced by<br />

diselenides. per-Sec <strong>peptides</strong> have been obtained by<br />

recombinant techniques (114-116) as well as by<br />

synthesis. Using Bzl(4-Me) for the selenol protection<br />

instead <strong>of</strong> the Mob group, which was applied in the<br />

synthesis <strong>of</strong> apamin analogues (99), and<br />

correspondingly the Boc chemistry, β-elimination at<br />

the Sec(Mob) residues was avoided and αselenoconotoxins<br />

were efficiently synthesized. These<br />

retained full bioactivities <strong>of</strong> the wild-type toxin and<br />

showed the expected high stability under reducing<br />

biological media where the wild-type toxin was fully<br />

deactivated (117).<br />

REFERENCES AND NOTES<br />

1. I. Annis, B. Hargittai, G. Barany, in: Methods Enzymol., 289<br />

198-221 (1997)<br />

2. L. Moroder, D. Besse, H. J. Musiol, S. Rudolph-Böhner, F. Siedler;<br />

Biopolymers 40 207-234 (1996)<br />

3. K. Akaji, Y. Kiso, in: Houben-Weyl, Methods <strong>of</strong> Organic Chemistry,<br />

<strong>Synthesis</strong> <strong>of</strong> <strong>Peptides</strong> and Peptidomimetics Vol. E22b, M. Goodman,<br />

A. Felix, L. Moroder, C. Toniolo, eds., Thieme, Stuttgart, 2002, pp.<br />

101-141<br />

4. L. Moroder, H.-J. Musiol, N. Schaschke, L. Chen, B. Hargittai, G.<br />

Barany, in: Houben-Weyl, Methods <strong>of</strong> Organic Chemistry,<br />

<strong>Synthesis</strong> <strong>of</strong> <strong>Peptides</strong> and Peptidomimetics, Vol. E22a, M.<br />

Goodman, A. Felix, L. Moroder, C. Toniolo, ed., Thieme,<br />

Stuttgart, 2003, pp. 384-423<br />

5. L. Moroder, H.-J. Musiol, M. Götz, C. Renner; Biopolymers 80 85-97<br />

(2005)<br />

6. R. Eritja, J. P. Ziehlermartin, P. A. Walker, T. D. Lee, K. Legesse, F.<br />

Albericio, B. E. Kaplan; Tetrahedron 43 2675-2680 (1987)<br />

7. E. Atherton, N. L. Benoiton, E. Brown, R. C. Sheppard, B. J. Williams;<br />

J, Chem. Soc. - Chem. Commun. 336-337 (1981)<br />

8. Y. Fujiwara, K. Akaji, Y. Kiso; Chem. Pharm. Bull. Jpn 42 7<strong>24</strong>-726<br />

(1994)<br />

9. T. Kaiser, G. J. Nicholson, H. J. Kohlbau, W. Voelter; Tetrahedron Lett.<br />

37 1187-1190 (1996)<br />

10. J. Lukszo, D. Patterson, F. Albericio, S. A. Kates; Lett. Peptide Sci. 3<br />

157-166 (1996)<br />

11. H. J. Musiol, F. Siedler, D. Quarzago, L. Moroder; Biopolymers 34<br />

1553-1562 (1994)<br />

12. F. Siedler, E. Weyher, L. Moroder; J. Peptide Sci. 2 271-275 (1996)<br />

13. B. H. Rietman, R. Smulders, I. F. Eggen, A. Vanvliet, G. Vandewerken,<br />

G. I. Tesser; Int. J. Peptide Protein Res. 44 199-206 (1994)<br />

14. G. Barany, Y. X. Han, B. Hargittai, R. Q. Liu, J. T. Varkey;<br />

Biopolymers 71 652-666 (2003)<br />

15. Y. T. Kung, Y. C. Du, W. T. Huang, C. C. Chen, L. T. Ke, S. C. Hu, R.<br />

Q. Jiang, S. Q. Chu, C. I. Niu, J. Z. Hsu, W. C. Chang, L. L. Cheng,<br />

H. S. Li, Y. Wang, T. P. Loh, A. H. Chi, C. H. Li, P. T. Shi, Y. H. Yieh, K.<br />

L. Tang, C. Y. Hsing; Sci. Sinica 14 1710-1716 (1965)<br />

16. P. G. Katsoyannis; Science 154 1509-1514 (1966)<br />

17. P. G. Katsoyannis, A. Tometsko; Proc. Natl. Acad. Sci. USA 55 1554-<br />

1561 (1966)<br />

18. J. Meienh<strong>of</strong>er, E. Schnabel, H. Bremer, O. Brinkh<strong>of</strong>f, R. Zabel, W.<br />

Sroka, T. H. Klostermeyer, D. Brandenburg, T. Okuda, H. Zahn; Z.<br />

Naturforsch. B 18 1120-1121 (1963)<br />

19. P. Sieber, B. Kamber, A. Hartmann, A. Jöhl, B. Riniker, W. Rittel; Helv.<br />

Chim. Acta 60 27-37 (1977)<br />

20. P. Sieber, B. Kamber, A. Hartmann, A. Jöhl, B. Riniker, W. Rittel; Helv.<br />

Chim. Acta 57 2617-2621 (1974)<br />

21. K. Maruyama, K. Nagata, M. Tanaka, H. Nagasawa, A. Isogai, H.<br />

Ishizaki, A. Suzuki; J. Protein Chem. 11 1-12 (1992)<br />

22. K. Akaji, K. Fujino, T. Tatsumi, Y. Kiso; J. Am. Chem. Soc. 115<br />

11384-11392 (1993)<br />

23. F. Lin, L. Otvos, J. Kumagai, G. W. Tregear, R. A. D. Bathgate, J. D.<br />

Wade; J. Peptide Sci. 10 257-264 (2004)<br />

<strong>24</strong>. E. E. Büllesbach, C. Schwabe; J. Biol. Chem. 266 10754-10761<br />

(1991)<br />

25. R. A. D. Bathgate, C. S. Samuel, T. C. D. Burazin, S. Layfield, A. A.<br />

Claasz, I. G. T. Reytomas, N. F. Dawson, C. X. Zhao, C. Bond, R. J.<br />

Summers, L. J. Parry, J. D. Wade, G. W. Tregear; J. Biol. Chem. 277<br />

1148-1157 (2002)<br />

26. R. A. D. Bathgate, F. Lin, N. F. Hanson, L. Otvos, A. Guidolin, C.<br />

Giannakis, S. Bastiras, S. L. Layfield, T. Ferraro, S. Ma, C. X. Zhao,<br />

A. L. Gundlach, C. S. Samuel, G. W. Tregear, J. D. Wade;<br />

Biochemistry 45 1043-1053 (2006)<br />

27. T. Koide, A. Otaka, H. Suzuki, N. Fujii; Synlett 345-346 (1991)<br />

28. K. Akaji, T. Tatsumi, M. Yoshida, T. Kimura, Y. Fujiwara, Y. Kiso; J.<br />

Am. Chem. Soc. 114 4137-4143 (1992)<br />

29. P. Sieber, B. Kamber, B. Riniker, W. Rittel; Helv. Chim. Acta 63 2358-<br />

2363 (1980)<br />

30. M. Vila-Perello, S. Tognon, A. Sanchez-Vallet, F. Garcia-Olmedo, A.<br />

Molina, D. Andreu; J. Med. Chem. 49 448-451 (2006)<br />

31. M. Vila-Perello, A. Sanchez-Vallet, F. Garcia-Olmedo, A. Molina, D.<br />

Andreu; J. Biol. Chem. 280 1661-1668 (2005)<br />

32. E. Klüver, K. Adermann, A. Schulz; J. Peptide Sci. 12 <strong>24</strong>3-257<br />

(2006)<br />

33. E. Klüver, S. Schulz-Maronde, S. Scheid, B. Meyer, W. G. Forssmann,<br />

K. Adermann; Biochemistry 44 9804-9816 (2005)<br />

34. A. Schulz, E. Klüver, S. Schulz-Maronde, K. Adermann; Biopolymers<br />

80 34-49 (2005)<br />

35. A. Otaka, T. Koide, A. Shide, N. Fujii; Tetrahedron Lett. 32 1223-<br />

1226 (1991)<br />

36. T. J. Wallace, J. J. Mahon; J. Org. Chem. 30 1502-1506 (1965)<br />

37. A. Cuthbertson, B. Indrevoll; Tetrahedron Lett. 41 3661-3663 (2000)<br />

38. A. Cuthbertson, E. Jarnaess, S. Pullan; Tetrahedron Lett. 42 9257-<br />

9259 (2001)<br />

39. B. Kamber, A. Hartmann, K. Eisler, B. Riniker, H. Rink, P. Sieber, W.<br />

Rittel; Helv. Chim. Acta 63 899-915 (1980)<br />

<strong>Peptides</strong><br />

chimica oggi • Chemistry Today • Vol <strong>24</strong> nr 4 • July/August 2006 35


<strong>Peptides</strong><br />

36<br />

40. C. Klinguer-Hamour, M. C. Bussat, H. Plotnicky, D. Velin, N. Corvaia,<br />

T. Nguyen, A. Beck; J. Peptide Res. 62 27-36 (2003)<br />

41. T. A. Lyle, S. F. Brady, T. M. Ciccarone, C. D. Colton, W. J. Paleveda,<br />

D. F. Veber, R. F. Nutt; J. Org. Chem. 52 3752-3759 (1987)<br />

42. J. P. Tam, C. R. Wu, W. Liu, J. W. Zhang; J. Am. Chem. Soc. 113<br />

6657-6662 (1991)<br />

43. I. Annis, L. Chen, G. Barany; J. Am. Chem. Soc. 120 7226-7238<br />

(1998)<br />

44. K. Darlak, D. W. Long, A. Czerwinski, M. Darlak, F. Valenzuela, A. F.<br />

Spatola, G. Barany; J. Peptide Res. 63 303-312 (2004)<br />

45. K. Maruyama, H. Nagasawa, A. Suzuki; <strong>Peptides</strong> 20 881-884<br />

(1999)<br />

46. A. Cuthbertson, B. Indrevoll; Org. Lett. 5 2955-2957 (2003)<br />

47. J. Ottl, L. Moroder; Tetrahedron Lett. 40 1487-1490 (1999)<br />

48. J. M. Fletcher, R. A. Hughes; Tetrahedron Lett. 45 6999-7001 (2004)<br />

49. J. Ottl, L. Moroder; J. Am. Chem. Soc. 121 653-661 (1999)<br />

50. T. Koide, Y. Nishikawa, Y. Takahara; Bioorg. Med. Chem. Lett. 14<br />

125-128 (2004)<br />

51. D. Barth, H.-J. Musiol, M. Schütt, S. Fiori, A. G. Milbradt, C. Renner,<br />

L. Moroder; Chem. Eur. J. 9 3692-3702 (2003)<br />

52. B. Saccà, L. Moroder; J. Peptide Sci. 8 192-204 (2002)<br />

53. C. J. Armishaw, P. F. Alewood; Curr. Protein Peptide Sci. 6 221-<strong>24</strong>0<br />

(2005)<br />

54. Z. B. Wu, D. M. Hoover, D. Yang, C. Boulègue, F. Santamaria, J. J.<br />

Oppenheim, J. Lubkowski, W. Y. Lu; Proc. Natl. Acad. Sci. USA 100<br />

8880-8885 (2003)<br />

55. C. J. Camacho, D. Thirumalai; Proc. Natl. Acad. Sci. USA 92 1277-<br />

1281 (1995)<br />

56. S. Goulas, D. Gatos, K. Barlos; J. Peptide Sci. 12 116-123 (2006)<br />

57. J. P. Tam, Y. A. Lu; Tetrahedron Lett. 38 5599-5602 (1997)<br />

58. J. C. Spetzler, C. Rao, J. P. Tam; Int. J. Peptide Protein Res. 43 351-<br />

358 (1994)<br />

59. J. Y. Chang; Biochem. J. 300 643-650 (1994)<br />

60. B. Chatrenet, J. Y. Chang; J. Biol. Chem. 268 20988-20996 (1993)<br />

61. B. Chatrenet, J. Y. Chang; J. Biol. Chem. 267 3038-3043 (1992)<br />

62. C. B. Anfinsen; Science 181 223-230 (1973)<br />

63. T. Kimura, in: Houben-Weyl, Methods <strong>of</strong> Organic Chemistry,<br />

<strong>Synthesis</strong> <strong>of</strong> <strong>Peptides</strong> and Peptidomimetics, Vol. E22b, M. Goodman,<br />

A. Felix, L. Moroder, C. Toniolo, eds., Thieme, Stuttgart, 2003, pp.<br />

142-161<br />

64. D. J. Craik, N. L. Daly, T. Bond, C. Waine; J. Mol. Biol. 294 1327-<br />

1336 (1999)<br />

65. Y. Q. Tang, J. Yuan, G. Osapay, K. Osapay, D. Tran, C. J. Miller, A.<br />

J. Ouellette, M. E. Selsted; Science 286 498-502 (1999)<br />

66. S. Luckett, R. S. Garcia, J. J. Barker, A. V. Konarev, P. R. Shewry, A.<br />

R. Clarke, R. L. Brady; J. Mol. Biol. 290 525-533 (1999)<br />

67. Y. Q. Qian, C. L. Tsou; Biochem. Biophys. Res. Commun. 146 437-<br />

442 (1987)<br />

68. J. G. Tang, C. L. Tsou; Biochem. J. 268 429-435 (1990)<br />

69. J. G. Tang, C. C. Wang, C. L. Tsou; Biochem. J. 255 451-455 (1988)<br />

70. J. D. Wade, G. W. Tregear, Methods Enzymol. 289 637-646 (1997)<br />

71. J. G. Tang, Z. H. Wang, G. W. Tregear, J. D. Wade; Biochemistry 42<br />

2731-2739 (2003)<br />

72. S. R. Woodward, L. J. Cruz, B. M. Olivera, D. R. Hillyard; EMBO J. 9<br />

1015-1020 (1990)<br />

73. W. D. Fairlie, H. P. Zhang, W. M. Wu, S. L. Pankhurst, A. R. Bauskin,<br />

P. K. Russell, P. K. Brown, S. N. Breit; J. Biol. Chem. 276 16911-<br />

16918 (2001)<br />

74. A. Rattenholl, M. Ruoppolo, A. Flagiello, M. Monti, F. Vinci, G.<br />

Marino, H. Lilie, E. Schwarz, R. Rudolph; J. Mol. Biol. 305 523-533<br />

(2001)<br />

75. Y. Hidaka, C. Shimono, M. Ohno, N. Okumura, K. Adermann, W.<br />

G. Forssmann, Y. Shimonishi; J. Biol. Chem. 275 25155-25162<br />

(2000)<br />

76. Y. Hidaka, M. Ohno, B. Hemmasi, O. Hill, W. G. Forssmann, Y.<br />

Shimonishi; Biochemistry 37 8498-8507 (1998)<br />

77. M. Price-Carter, W. R. Gray, D. P. Goldenberg; Biochemistry 35<br />

15537-15546 (1996)<br />

78. M. Price-Carter, W. R. Gray, D. P. Goldenberg; Biochemistry 35<br />

15547-15557 (1996)<br />

79. J. S. Weissman, P. S. Kim; Cell 71 841-851 (1992)<br />

80. A. Rattenholl, H. Lilie, A. Grossmann, A. Stern, E. Schwarz, R.<br />

Rudolph; Eur. J. Biochem. 268 3296-3303 (2001)<br />

81. K. Nakajima, S. Kubo, S. Kumagaye, H. Nishio, M. Tsunemi, T. Inui,<br />

H. Kuroda, N. Chino, T. X. Watanabe, T. Kimura, S. Sakakibara;<br />

Biochem. Biophys. Res. Commun. 163 4<strong>24</strong>-429 (1989)<br />

82. E. L. Cunningham, T. Mau, S. M. E. Truhlar, D. A. Agard;<br />

Biochemistry 41 8860-8867 (2002)<br />

83. O. Buczek, B. M. Olivera, G. Bulaj; Biochemistry 43 1093-1101<br />

(2004)<br />

84. G. Bulaj, O. Buczek, I. Goodsell, E. C. Jimenez, J. Kranski, J. S.<br />

Nielsen, J. E. Garrett, B. M. Olivera; Proc. Natl. Acad. Sci. USA 100<br />

14562-14568 (2003)<br />

85. P. Buczek, O. Buczek, G. Bulaj; Biopolymers 80 50-57 (2005)<br />

86. T. S. Kang, S. Vivekanandan, S. D. S. Jois, R. M. Kini; Angew. Chem.<br />

Int. Ed. 44 6333-6337 (2005)<br />

87. H. Tamaoki, R. Miura, M. Kusunoki, Y. Kyogoku, Y. Kobayashi, L.<br />

Moroder; Protein Engin. 11 649-659 (1998)<br />

88. P. M. Harrison, M. J. E. Sternberg; J. Mol. Biol. 264 603-623 (1996)<br />

89. L. Narasimhan, J. Singh, C. Humblet, K. Guruprasad, T. Blundell;<br />

Nature Struct. Biol. 1 850-852 (1994)<br />

90. R. S. Norton, P. K. Pallaghy; Toxicon 36 1573-1583 (1998)<br />

91. P. K. Pallaghy, K. J. Nielsen, D. J. Craik, R. S. Norton; Protein Sci. 3<br />

1833-1839 (1994)<br />

92. A. Gupta, H. W. T. Van Vlijmen, J. Singh; Protein Sci. 13 2045-2058<br />

(2004)<br />

93. C. C. Chuang, C. Y. Chen, J. M. Yang, P. C. Lyu, J. K. Hwang; Protein<br />

Struct. Funct. Genet. 53 1-5 (2003)<br />

94. S. Cheek, S. S. Krishna, N. V. Grishin; J. Mol. Biol. 359 215-237<br />

(2006)<br />

95. B. M. P. Huyghues-Despointes, J. W. Nelson; Biochemistry 31 1476-<br />

1483 (1992)<br />

96. S. I. Kumagaye, H. Kuroda, K. Nakajima, T. X. Watanabe, T. Kimura,<br />

T. Masaki, S. Sakakibara; Int. J. Peptide Protein Res. 32 519-526<br />

(1988)<br />

97. R. M. Zhang, G. H. Snyder; J. Biol. Chem. 264 18472-18479<br />

(1989)<br />

98. D. Besse, F. Siedler, T. Diercks, H. Kessler, L. Moroder; Angew. Chem.<br />

Int. Ed. Engl. 36 883-885 (1997)<br />

99. S. Fiori, S. Pegoraro, S. Rudolph-Böhner, J. Cramer, L. Moroder;<br />

Biopolymers 53 550-564 (2000)<br />

100.S. Pegoraro, S. Fiori, J. Cramer, S. Rudolph-Böhner, L. Moroder;<br />

Protein Sci. 8 1605-1613 (1999)<br />

101.D. Besse, N. Budisa, W. Karnbrock, C. Minks, H. J. Musiol, S.<br />

Pegoraro, F. Siedler, E. Weyher, L. Moroder; Biol. Chem. 378 211-<br />

218 (1997)<br />

102.L. Moroder; J. Peptide Sci. 11 187-214 (2005)<br />

103.T. Shi, S. M. Spain, D. L. Rabenstein; Angew. Chem. Int. Ed. 45<br />

1780-1783 (2006)<br />

104.E. M. Kurz, T. W. Holstein, B. M. Petri, J. Engel, C. N. David; J. Cell<br />

Biol. 115 1159-1169 (1991)<br />

105.E. Pokidysheva, A. G. Milbradt, S. Meier, C. Renner, D. Haussinger,<br />

H. P. Bächinger, L. Moroder, S. Grzesiek, T. W. Holstein, S. Özbek, J.<br />

Engel; J. Biol. Chem. 279 30395-30401 (2004)<br />

106.A. G. Milbradt, C. Boulègue, L. Moroder, C. Renner; J. Mol. Biol. 354<br />

591-600 (2005)<br />

107.C. Boulègue, A. G. Milbradt, C. Renner, L. Moroder; J. Mol. Biol. 358<br />

846-856 (2006)<br />

108.J. D. Gough, J. M. Gargano, A. E. Don<strong>of</strong>rio, W. J. Lees; Biochemistry<br />

42 11787-11797 (2003)<br />

109.J. D. Gough, R. H. Williams, A. E. Don<strong>of</strong>rio, W. J. Lees; J. Am.<br />

Chem. Soc. 1<strong>24</strong> 3885-3892 (2002)<br />

110.K. J. Woycechowsky, K. D. Wittrup, R. T. Raines; Chem. Biol. 6 871-<br />

879 (1999)<br />

111.R. DeLa Cruz, F. G. Whitby, O. Buczek, G. Bulaj; J. Peptide Res. 61<br />

202-212 (2003)<br />

112.L. Martin, P. Barthe, O. Combes, C. Roumestand, C. Vita; Tetrahedron<br />

56, 9451-9460 (2000)<br />

113.A. J. Nicoll, C. J. Weston, C. Cureton, C. Ludwig, F. Dancea, N.<br />

Spencer, O. S. Smart, U. L. Günther, R. K. Allemann; Org. Biomol.<br />

Chem. 3, 4310-4315 (2005)<br />

114.S. Müller, H. Senn, B. Gsell, W. Vetter, C. Baron, A. Böck;<br />

Biochemistry 33, 3404-3412 (1994)<br />

115.J.-F. Sanchez, F. Hoh, M.-P. Strub, A. Aumelas, C. Dumas; Structure<br />

10, 1363-1370 (2002)<br />

116.M.-P. Strub, F. Hoh, J.-F. Sanchez, J. M. Strub, A. Böck, A. Aumelas,<br />

C. Dumas; Structure 11, 1359-1367 (2003)<br />

117.C. J. Armishaw, N. L. Daly, S. T. Nevin, D. J. Adams, D. J. Craik, P. F.<br />

Alewood; J. Biol. Chem. 281, 14136-14143 (2006)<br />

CYRIL BOULÈGUE, HANS-JÜRGEN MUSIOL,<br />

VIDYA PRASAD AND LUIS MORODER*<br />

*Corresponding author<br />

Max-Planck-Institut für Biochemie<br />

Am Klopferspitz 18,<br />

D-82152 Martinsried, Germany<br />

chimica oggi • Chemistry Today • Vol <strong>24</strong> nr 4 • July/August 2006

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!