Oxathiocoraline: Lessons to be learned from the synthesis of complex N-methylated depsipeptides

FULL PAPER

DOI: 10.1002/ejoc.200900259

Oxathiocoraline: Lessons to be Learned from the Synthesis of Complex

N-Methylated Depsipeptides

Núria Bayó-Puxan,^[a]Judit Tulla-Puche,*^[a,b]and Fernando Albericio*^[a,b,c]

Keywords: Thiocoraline / Peptides / Solid-phase synthesis / Cysteine / Diketopiperazines / Synthesis design

Investigations into the synthesis of oxathiocoraline, a bicyclic

depsipeptide with C₂symmetry, revealed a number of unex-

pected side-reactions that could not be circumvented by clas-

sical or standard means. This cyclodepsipeptide has a large

number of N-methyl amino acids coexisting with two ester

bonds and also shows a branched structure; these features

hinder its synthesis. In addition, complexity is further in-

creased because of the presence of a large number of non-

commercially available cysteines and heterocyclic moieties.

Here we describe the general points that should be ad-

dressed when attempting the synthesis of a cyclodepsipep-

tide, such as strategies to prevent or minimize diketopiperaz-

ine formation, β-elimination, and oxidation byproducts. The

optimum design includes a suitable solid support, the choice

of the best starting point for the synthesis (first amino acid to

anchor to the resin) and a set of compatible and/or protecting

groups and coupling reagents.

Germany, 2009)

methyl amino acids coexisting with two ester bonds is a

bottleneck because of side-reactions associated with these

Introduction

Cyclic depsipeptides of marine and terrestrial origin are two features.

frequently found to be novel potent compounds.^[1–3]The

development of the chemical synthesis of these compounds

is required in order to verify their structures and to provide

enough material for preclinical studies. Moreover, control

of their syntheses allows researchers to examine the poten-

tial to produce new analogues with improved pharmaceuti-

cal and pharmacokinetic properties. Given the complex

structures of this class of compounds, their synthesis im-

plies the fine-tuning of standard peptide chemistry pro-

cedures. In this regard, the ester bonds present in depsipep-

Figure 1. Structure and composition of oxathiocoraline.

tides present a double drawback: low reactivity of alcohols

in the coupling reactions and fragile break points in the

synthetic approach.

In addition, the synthetic difficulty is further increased

as a result of its branched structure and the presence of a

large number of non-commercially available cysteines and

heterocyclic moieties. Oxathiocoraline thus represents a

challenge that requires the re-evaluation of standard pep-

tide procedures. Recently, the first synthesis of oxathiocor-

aline has been achieved by an optimized solid-phase ap-

proach, through a key intermolecular disulfide bond forma-

tion on solid-support.^[4]Together with its amide ana-

Oxathiocoraline^[4]is a bicyclic depsipeptide with C₂sym-

metry. It is made up of two antiparallel tetrapeptide chains

linked by two ester bonds and an additional disulfide bridge

in the central part of the molecule (Figure 1). From a syn-

thetic point of view, the presence of a large number of N-

[a] Institute for Research in Biomedicine, Barcelona Science Park, logue,^[5,6]oxathiocoraline was designed with the aim of pro-

Baldiri Reixac 10, 08028 Barcelona, Spain

ducing a more resistant compound displaying antitumour

Fax: +34-93-403-71-26

activity comparable to that of the natural compound thio-

coraline.^[7,8]Here we discuss and analyse the procedures

followed to achieve the synthesis of oxathiocoraline. Our

strategy was based on the following steps: firstly, the design

and evaluation of the possible synthetic approaches, taking

into account the C₂symmetry of the molecule and the dis-

tinct starting points, and secondly, an exhaustive explora-

E-mail: judit.tulla@irbbarcelona.org

albericio@irbbarcelona.org

[b] CIBER-BBN, Networking Centre on Bioengineering, Biomate-

rials and Nanomedicine, Barcelona Science Park

Baldiri Reixac 10, 08028 Barcelona, Spain

[c] Department of Organic Chemistry, University of Barcelona

Martí i Franquès 1–11, 08028 Barcelona, Spain

Supporting information for this article is available on the

WWW under http://dx.doi.org/10.1002/ejoc.200900259.

Eur. J. Org. Chem. 2009, 2957–2974

2957

N. Bayó-Puxan, J. Tulla-Puche, F. Albericio

FULL PAPER

tion of the most feasible synthetic approaches, paying spe-

cial attention to the most relevant side reactions: diketopip-

erazine (DKP) formation, ester stability and Cys-based

side-reactions. Valuable information from the side-reactions

encountered should be usable to establish a set of guidelines

for subsequent research in cyclic depsipeptide chemistry.

Results and Discussion

Scheme 1. Possible strategies for the construction of a symmetrical

peptide: A) stepwise approach, B) 4+4 fragment coupling, or

C) 4+4 fragment dimerization.

1. Analysis of Possible Synthetic Approaches

1.1. Symmetry in the Synthetic Peptide Design

1.2. Choice of the Amino Acid Attached to the Solid

Support

By consideration of the symmetry of oxathiocoraline, the

number of synthetic steps was reduced, and the number of

side reactions minimized. Oxathiocoraline should thus be

Given the bicyclic structure of oxathiocoraline, the first

better constructed by fragment coupling (Scheme 1, B) or amino acid will be involved in three crucial steps: in correct

by fragment dimerization (Scheme 1, C) than by a stepwise peptide chain elongation, in fragment coupling and in the

approach (Scheme 1, A). In the first two cases, only one es- final cyclization as the C-terminal amino acid. In a first

ter bond within the peptide chain would need to be con- analysis, none of the four amino acids that make up the

structed, because the second one would be circumvented by oxathiocoraline can either be discounted or chosen as an

fragment coupling. All the drawbacks associated with the obvious starting point, and an accurate evaluation of each

presence of ester bonds would therefore be reduced.

is required. Each possibility must be analysed in terms of

Figure 2. Graphic representation of potential side-reactions for each starting point in oxathiocoraline synthesis. ᭹ indicates the ester

bond; ᭡ indicates the fragment coupling bond. * DKP formation will take place only when a total stepwise elongation is carried out; if

a convergent approach is used this DKP formation does not occur.

2958

www.eurjoc.org

Eur. J. Org. Chem. 2009, 2957–2974

Synthesis of Complex N-Methylated Depsipeptides

the risk of DKP formation, racemization and – when Cys - and -amino acids, as well as the presence of one N-

is the C-terminal amino acid – racemization and β-elimi- methylamino acid – or, worse still, two N-methylamino

nation, as well as difficulties both in fragment coupling and acids – favour DKP formation and must be treated care-

in cyclization. Figure 2 illustrates the four starting points fully. A number of general precautions have been described,

and their associated risks.

such as the use of highly steric hindered resins (e.g., CTC

resin, which protects the carboxylic function^[14–16]) or pro-

tecting groups that are removed under acidic or almost neu-

tral conditions.^[17–20]In depsipeptide syntheses, the presence

of labile ester bonds in the peptide chain provides another

potential point for DKP formation. This process is fav-

oured by the flexibility of the growing chain together with

the absence of protection from the solid-support steric hin-

drance. A general approach to achieving long peptides or

complex sequences is the use of the fragment coupling ap-

proach. In this case, the active ester of the fragment to be

coupled to the peptidyl resin is formed in situ in solution.

However, this active fragment can also undergo DKP for-

mation under specific conditions, as we encountered in this

synthetic exploration. DKP formation can therefore occur

at the resin level, at the intrachain level and at the in-

terchain fragment activation level.

NMe-Cys(Me)

Disadvantages: Cys as C-terminal amino acid is associ-

ated with racemization and β-elimination as side-reac-

tions.^[9–12]Moreover, two NMe-amino acids at the C ter-

minus are highly prone to give DKP formation.^[13]Advan-

tages: Having -Ser at the last position allows the introduc-

tion of the heterocycle after elongation of the tetrapeptidic

chain. A further advantage is that the ester bond, which is

highly labile, is constructed in the last step.

NMe-Cys(Acm)

Disadvantages: The same problems as those associated

with the presence of a Cys at the C terminus. Moreover, a

fragment coupling approach or a cyclization requires cou-

pling between two NMe-amino acids. Advantages: There

should be slightly less DKP formation because there are no

consecutive N-methyl amino acids, although the Gly-NMe-

AA sequence is also prone to give DKP.

a) At the Resin Level

When Cys(Me) is the first amino acid [Figure 2, NMe-

Cys(Me)], DKP formation can occur on deprotection of the

dipeptide, because the sequence NMe-Cys(Acm)-NMe-

Cys(Me) (two consecutive NMe-AAs) is highly prone to

give DKPs. The formation of the tetrapeptides was thus

first examined with use of non-methylated Cys, both on

Wang resin and on CTC resin.^[21]Two strategies were used

(Scheme 2). For Wang resin, Alloc-Cys(Me)-OH (Scheme 2,

a) was introduced as the first amino acid, and the pNZ

group was used for protection of the amino group of the

second residue, Cys(Acm), to keep the amino group proton-

ated and therefore to minimize DKP formation.^[20]Alloc-

Gly-OH was introduced as a third amino acid, and finally

Fmoc--Ser(Trt)-OH completed the tetrapeptide. In this

approach, cleavage of the Fmoc group was used to quantify

the resin loading. For CTC resin (Scheme 2, b), the tripep-

tide was elongated through Alloc chemistry, because pNZ

acidic removal conditions result in partial detachment of

the peptidic chain. As in the previous strategy, Fmoc--

Ser(Trt)-OH was also used as fourth amino acid.

For the synthesis with Wang resin, Fmoc quantitation

showed 100% DKP formation, thereby confirming that

DKPs form very rapidly and that protonation of the amino

group is not sufficient to prevent this process. A high degree

of DKP formation was also observed for the synthesis with

CTC resin, although it was minimized by an optimized

Neutral Removal Coupling Step: the rapid removal of the

Alloc group (2ϫ10 min) followed by two consecutive fast

couplings of Alloc-Gly-OH with HATU/HOAt with fewer

equiv. of base. The most reactive aminium salt, HATU

(5 equiv.), which is specially indicated for NMe-amino acid

Gly

Disadvantages: The main limitation of this approach is

DKP formation, which can occur in the two key steps: 4+4

fragment coupling and cyclization. Advantages: No racemi-

zation will occur when loading the amino acid on the resin

or in a 4+4 fragment coupling or cyclization. No DKP for-

mation will occur at the resin level because the second resi-

due (-Ser) is elongated through its side chain.

D

-Ser

Disadvantages: The ester bond is constructed at a very

early stage in the sequence and is therefore more exposed

to protecting group removal conditions that may break it.

Advantages: The fragment coupling and the cyclization step

should proceed smoothly because the two residues involved

are the non-methylated amino acids (-Ser and Gly).

Furthermore, because the depsipeptide is elongated through

the side chain, no DKP should be formed at the resin level.

Since none of the four possibilities is a priori exempt

from possible side-reactions, the four starting points were

explored in various synthetic stages: tetrapeptide synthesis,

octapeptide elongation, cyclization and heterocycle moiety

incorporation.

2. Difficulties and Successes in Peptide Elongation

2.1. DKP Formation Throughout the Sequence

DKP formation is a side-reaction frequently encountered coupling, was used, although with a smaller amount of base

at the unprotected dipeptidyl-resin level. This secondary re- (4.8 equiv.) than used in standard procedures (10–

action is highly sequence-dependent, although some trends 15 equiv.). In this case, we suspect that 4.8 equiv. of active

for its formation have been described.^[13]Combinations of ester was formed and that it reacted more rapidly with the

Eur. J. Org. Chem. 2009, 2957–2974

www.eurjoc.org

2959

N. Bayó-Puxan, J. Tulla-Puche, F. Albericio

FULL PAPER

Scheme 2. Study of DKP formation on Wang and CTC resin.

NMe-amino acid. To ensure completion of the coupling, a NMe-Cys(Acm)-NMe-Leu-OH by solid-phase N-methyl-

re-coupling was performed under the same conditions. A ation^[25]had provided the tripeptides in high yields. Solid-

negative De Clercq test^[22]indicated total conversion.

phase N-methylation, which requires a long period of time

When the synthesis was attempted with NMe-amino in basic media, was carried out on the dipeptidyl resin H-

acids and Wang resin, a further limitation was encountered: Cys(Acm)-NMe-Leu-O-CTC, and Alloc-Gly-OH was then

the acidic treatment (50% TFA in CH₂Cl₂) applied to re- introduced without any loss of yield (Scheme 3).

lease the protected dipeptide from the resin resulted in loss

of the NMe-amino acid at the C-terminal position. This

side-reaction had been described in the case of Boc chemis-

try.^[23,24]When CTC resin was used, the dipeptide Alloc-

Gly-NMe-Cys(Acm)-OH was coupled to H-NMe-Cys(Me)-

O-CTC. The coupling gave the desired tripeptide although

a high degree of racemization at the NMe-Cys(Acm) site

was observed; Cys at the C terminus is prone to give race-

mization with a further increase if the residue is N-acylated.

Finally, a stepwise approach was attempted by applica-

tion of Alloc chemistry and by minimization of reaction

time and the basicity. Shorter treatments for Alloc removal

were therefore applied (2ϫ10 min), and a deficit of DIEA

Scheme 3. Synthesis of the tripeptide Alloc-Gly-NMe-Cys(Acm)-

NMe-Leu-O-CTC resin with use of methylation on solid phase at

was used (0.8 equiv.) for the coupling step. The tripeptide

Alloc-Gly-NMe-Cys(Acm)-NMe-Cys(Me)-OH was ob-

tained as a major compound in 70% yield. It is well known

that DKP formation is highly sequence-dependent. Pre-

the second amino acid.

When NMe-Cys(Acm) was used as the first amino acid

vious assembly of the tripeptides Alloc-Gly-NMe-Leu- [Figure 2, NMe-Cys(Acm)], DKP formation occurred at the

NMe-Leu-OH by standard Fmoc chemistry and Alloc-Gly- Gly deprotection step. In this case, the use of the Alloc

2960

www.eurjoc.org

Eur. J. Org. Chem. 2009, 2957–2974

Synthesis of Complex N-Methylated Depsipeptides

group to protect the second amino acid, combined with pre-

activation of the third amino acid (Boc--Ser-OH), ensured

high yields of the desired tripeptide (Scheme 4).

Scheme 5. DKP formation by attack on the ester bond within the

peptide chain during depsipeptide synthesis.

Starting the tetrapeptide at Gly (Figure 2, Gly) prevents

DKP formation at the resin level. However, DKP formation

at the -Ser ester bond can occur during fragment coupling.

We tested two sets of coupling conditions: DIPCDI/HOAt

and PyOAP/DIEA. The neutral conditions were used to

minimize DKP formation and to preserve ester integrity

through the use of non-basic media, whereas the harsher

conditions were assayed to favour a rapid coupling reac-

tion.^[27]In all cases, coupling with DIPCDI/HOAt resulted

in the recovery of the unreacted tetrapeptide. For coupling

with PyOAP/DIEA, low levels of conversion were achieved

because of the basic conditions. Treatment with the pro-

tected Fmoc fragment resulted in a preferential intramolec-

ular attack (DKP formation), rather than the desired inter-

molecular coupling (Scheme 6). Analysis of the peptidyl

resin after a 24 h coupling period thus showed the presence

of the octapeptide but also the appearance of the dipeptide

Boc--Ser-Gly-OH. Analysis of the coupling solution

showed the presence of the DKP [&NMe-Cys(Acm)-NMe-

Scheme 4. Synthesis of tripeptide Boc--Ser-Gly-NMe-Cys(Acm)-

O-CTC resin.

Starting the synthesis at Gly (Figure 2, Gly) has the ad-

vantage that it prevents DKP formation at the resin level

because the elongation of the peptide chain takes place

through the side chain of -Ser. The tetrapeptide [H-NMe-

Cys(Acm)-NMe-Cys(Me)&][Boc--Ser(&)-Gly-O-CTC]^[26]

was thus obtained in good yield and purity. However, the

problem is that DKPs can form through the Ser ester bond,

as further explained in the next section.

b) At the Ester Level in the Peptide Chain

The synthesis of depsipeptides involves the presence of Cys(Me)&].

an ester bond in the sequence. In many synthetic ap-

When the synthesis was started at -Ser (Figure 2, -Ser)

proaches, this bond is located within the growing peptide and Alloc chemistry was used, only a 60% yield was ob-

chain and there is a risk of DKP formation after deprotec- tained for the tetrapeptide {[Alloc-Gly-NMe-Cys(Acm)-

tion of the second amino acid after the ester bond NMe-Cys(Me)&][Boc--Ser(&)-O-CTC]}. The byproducts

(Scheme 5). The growing peptide chain is now more flexible indicated DKP formation through the -Ser ester bond. To

and the effect of hindrance protection from the solid sup- suppress DKP formation, we used the tandem deprotection/

port is no longer noticeable.

coupling approach described by Thieriet et al.^[19]This ap-

Scheme 6. Fragment coupling products starting at Gly.

Eur. J. Org. Chem. 2009, 2957–2974

www.eurjoc.org

2961

N. Bayó-Puxan, J. Tulla-Puche, F. Albericio

FULL PAPER

proach uses an Alloc removal cocktail in which the incom-

ing Fmoc-protected amino acid is included in fluoride-acti-

vated form (Scheme 7).

Scheme 7. Deprotection/coupling tandem reaction described by

Thieriet et al. (Ref. 19).

However, the use of this approach was incompatible with

CTC resin. The peptidyl resin bond was unstable to the HF

generated from the tandem reaction and the incoming acti-

vated Fmoc-Gly-F ended up attached to the resin, so only

Fmoc-Gly-OH was recovered at the cleavage step. A second

approach was assayed by use of the optimized Neutral Re-

moval Coupling step described earlier (Figure 3).

Figure 4. HPLC analysis and structures of linear octadepsipeptide

and the byproduct resulting from loss of DKP [&NMe-Leu-NMe-

Leu&]. HPLC conditions: gradient from 7:3 to 0:10 in 15 min (A:

H₂O with 0.045% TFA and B: ACN with 0.036% TFA).

c) At the Activated C-Terminal End

Although the evaluation of NMe-Cys(Acm) as a starting

point concluded that this approach is not associated with

DKP formation either at the resin or at the peptide chain

level, it was surprising to detect intramolecular DKP for-

mation at the amide bond level during fragment coupling.

The protected tetrapeptide underwent an intramolecular re-

action between the Gly amide bond and the activated car-

boxylic function at the C terminus. There was competition

between intermolecular (fragment coupling) and intramo-

lecular (DKP formation) attack, which resulted in a moder-

ate coupling reaction yield of 75% (Scheme 8).

Figure 3. HPLC analysis from Fmoc-tetrapeptidyl resin starting at

-Ser. HPLC conditions: gradient from 8:2 to 0:10 in 15 min (A:

H₂O with 0.045% TFA and B: ACN with 0.036% TFA).

2.2. Protection Scheme and Side-Reactions Associated with

Ester Stability

Depsipeptide synthesis is associated with two main prob-

lems: reduced reactivity of the hydroxy group in the forma-

tion of the ester bond and at the same time high lability of

this newly formed bond. In the process of developing the

synthesis of oxathiocoraline, we explored the dependence of

these two key points on sequence and on protecting group

chemistry. Several side-reactions were identified.

We performed another synthesis using Alloc-NMe-Leu-

OH instead of both Alloc-NMe-Cys(Me)-OH and Alloc-

NMe-Cys(Acm)-OH. This approach revealed the effect of

chain flexibility. The first tetrapeptide {[Alloc-Gly-NMe-

Leu-NMe-Leu&][Boc--Ser(&)-O-CTC]} was obtained as a

pure compound, thereby indicating the absence of DKP

formation. After elongation of the peptide chain by a step-

wise approach, it was found that the second ester was more

labile, and the final octadepsipeptide was achieved in 50%

a) β-Elimination at the C Terminus

Acidity of the α-proton of an amino acid residue in com-

yield, with the other 50% corresponding to the hexapeptide bination with the presence of a leaving group at the β-posi-

arising from the loss of the DKP (Figure 4). In this case the tion favours a β-elimination reaction. Nature has translated

second ester bond showed more flexibility and so more this into the formation of didehydroalanine from Ser and

DKP was formed.

Cys residues, of the didehydro-α-aminobutyric acid moiety

2962

www.eurjoc.org

Eur. J. Org. Chem. 2009, 2957–2974

Synthesis of Complex N-Methylated Depsipeptides

Scheme 8. Fragment coupling products starting at NMe-Cys(Acm).

from Thr, and of didehydrophenylalanine from β-hydroxy-

phenylalanine.^[28–31]During the synthetic process, this β-eli-

mination can be enhanced by the presence of withdrawing

groups at the N-function (protecting groups).^[11,32]β-Elimi-

nation byproducts were detected when synthetic approaches

began with either -Ser or Cys residues. When a Fmoc-

based strategy was followed, further incorporation of piper-

idine through a Michael addition to give a 3-(piperidin-1-

yl)alanine residue was observed (Scheme 9).^[11]The extent

of this side-reaction was greater with further Fmoc removal

steps. When the peptide chain began with Fmoc-Cys(Acm)-

OH, analysis of the corresponding dipeptide H-Gly-Cys-

(Acm)-OH and tetrapeptide [Boc--Ser(&)-Gly-Cys(Acm)-

OH][H-Cys(Me)-&] showed the presence of the NMe-3-

(piperidin-1-yl)alanine residue instead of NMe-Cys(Acm),

increasing after each Fmoc treatment.

When the synthesis was repeated with use of a Fmoc pro-

tecting group only for the first amino acid, with Alloc

groups for Gly and NMe-Cys(Me), the tetrapeptide {[Boc-

-Ser(&)-Gly-NMe-Cys(Acm)-OH][Alloc-NMe-Cys(Me)&]}

was obtained free of NMe-3-(piperidin-1-yl)alanine. This

result indicated that this side-product was formed exclu-

sively when the amino function was N-acylated and not

during the removal of the Fmoc of the C-terminal Cys. The

Fmoc-protected group can thus be used for the first amino

acid, which is also useful for determination of the initial

functionalization, but Alloc chemistry should be used for

the following steps.

When the peptide chain began at -Ser, β-elimination

plus piperidine addition was also observed (Figure 5).

Analysis of protected and free dipeptide revealed quantita-

tive conversion into the compound QNA-3-(piperidin-1-yl)-

alanine. In this case, the combination of the electron-with-

drawing N-acyl form with the ester bond substantially in-

creased the acidity of the α-proton.

To prevent this side-reaction, a second attempt using a

carbamate-protecting group, such as the Boc group, was

made (Scheme 10).

Analysis of an aliquot of the final tetrapeptide revealed

the presence of the desired compound and the absence of

the 3-(piperidin-1-yl)alanine residue. However, another by-

product, resulting from the loss of the NMe-Leu residue,

was detected. This side-reaction was associated with the in-

stability of the ester bond inside the peptide chain, as de-

scribed in next section. Therefore, -Ser and Cys residues

at the C terminus are subject to two reactions: i) β-elimi-

Scheme 9. Mechanism of NMe-3-(piperidin-1-yl)alanine formation. nation with subsequent piperidine incorporation, and

Eur. J. Org. Chem. 2009, 2957–2974

www.eurjoc.org

2963

N. Bayó-Puxan, J. Tulla-Puche, F. Albericio

FULL PAPER

Figure 5. Synthesis of tetrapeptide starting at -Ser with coupling of the heterocycle on solid support. a) dipeptide {[QNA--Ser(&)-O-

CTC][Fmoc-NMe-Leu&]}, and b) compound obtained after piperidine/DMF (1:4) treatment. HPLC conditions: gradient from 7:3 to

0:10 in 15 min (A: H₂O with 0.045% TFA and B: ACN with 0.036% TFA).

Scheme 10. Synthesis of tetrapeptide starting at -Ser.

ii) loss of the adjacent unprotected amino acid on piper- these first attempts, either 3-hydroxyquinaldic acid (3-

idine treatment once the ester bond is formed.

HQA) or commercial quinaldic acid (QNA) were used

By using Alloc chemistry instead of Fmoc chemistry, we (Scheme 11).

obtained the tetradepsipeptides {[Alloc-Gly-NMe-Leu-

NMe-Leu&][Boc--Ser(&)-OH]} and {[Alloc-Gly-NMe-

Cys(Acm)-NMe-Cys(Me)&][Boc--Ser(&)-OH]} in high

yields and purities.

Table 1. Exploring diverse chemistries based on initiation at Gly to

afford the tetradepsipeptide.

R¹

R²

R³

Peptide Observations

purity

b) Ester Stability within the Peptide Chain

Natural cyclic depsipeptides generally incorporate a

branching point at a Ser/Thr residue, through the associated

hydroxy group. The stability and reactivity of the hydroxy

group can be affected by protecting group chemistry, cou-

pling conditions and the step at which the branching point

is created. When synthesis was started at Gly on Wang

resin, several protection strategies (Table 1) were examined.

The main limitation was found to reside in ester lability in

response to the protecting group removal conditions. In

1

2

3

4

Fmoc

QNA

Alloc pNZ good

pNZ pNZ poor

Alloc pNZ good

Alloc Boc good

Fmoc cannot be changed at a later

stage because of piperidine cleavage

cleavage of the ester with removal

of pNZ at position 2

upon deprotection of the tetrapep-

tide, partial cleavage of ester occurs

the unprotected tetradepsipeptide is

obtained

5

6

3-HQA Alloc pNZ poor

pNZ Alloc Boc good

elongation through the OH

2964

www.eurjoc.org

Eur. J. Org. Chem. 2009, 2957–2974

Synthesis of Complex N-Methylated Depsipeptides

Scheme 11. Synthesis of tetrapeptide starting at Gly.

Table 1 examines the purities of protected tetradepsipep- Cys(Acm) (Entry 1), a tetradepsipeptide of good purity was

tides obtained when starting at Gly. With use of a combina- obtained. pNZ was introduced at this position in order to

tion of Fmoc for -Ser, Alloc for Cys(Me) and pNZ for minimize DKP formation. However, this strategy has the

Scheme 12. Final strategies for tetradepsipeptides containing a) quinaldic acid, and b) 3-hydroxyquinaldic acid.

Eur. J. Org. Chem. 2009, 2957–2974

www.eurjoc.org

2965

N. Bayó-Puxan, J. Tulla-Puche, F. Albericio

FULL PAPER

drawback that Fmoc can not be removed after the ester has the dipeptide level, as in the previous strategy, because this

been formed because of cleavage of the ester bond upon quinolinic OH was found to be unreactive in solution reac-

treatment with piperidine. Switching to quinaldic acid after tions under mild coupling conditions (EDC·HCl/HOSu).

the introduction of Fmoc--Ser(Trt)-OH solves this prob- The introduction of the third and fourth amino acids, how-

lem. Elongation of the peptide by use of pNZ chemistry ever, resulted in additional elongation through the hydroxy

(Entry 2) gave a protected tetradepsipeptide of low purity function (Entry 5), and so it was necessary to change the

because of cleavage of the ester when SnCl₂was used to protection scheme in order to couple the heterocycle at a

remove the pNZ group from Cys(Me). When the Alloc later stage. With this in mind, the dipeptide was treated with

group was used instead of pNZ at this position, removal of piperidine to remove the Fmoc group, and the amino func-

the Alloc group with [Pd(PPh₃)₄] was clean, without cleav- tion was reprotected with pNZ-Cl. Ester formation was ac-

age of the ester functionality. pNZ-Cys(Acm)-OH was cou- complished as before, by the introduction of Alloc-

pled as a fourth amino acid, providing the tetradepsipeptide Cys(Me)-OH, but Boc was used as a protecting group for

in good purity (Entry 3). Although this strategy can be con- Cys(Acm). At this stage, the pNZ group was removed, and

tinued, partial cleavage of the ester bond occurred with the coupling with 3-hydroxyquinaldic acid was accomplished

SnCl₂treatment, probably because of DKP formation. This with DIPCDI/HOAt in DMF. The tetradepsipeptide was

strategy is therefore not appropriate when using NMe-AAs, obtained in good purities (Entry 6). The final strategies

because DKP formation is more favoured. A fourth strategy used to obtain the two tetradepsipeptides are thus summa-

using Boc-Cys(Acm)-OH was also examined. Cleavage of rized in Scheme 12. HPLC traces are shown in Figure 6.

the peptide from the resin simultaneously removed the Boc

c) TBDMS as a Convenient Protecting Group for Ser

group. In this case, the unprotected tetradepsipeptide was

obtained in good purity.

In the synthesis of cyclic depsipeptides, Ser is commonly

To obtain the natural target 3-hydroxyquinaldic-pro- used as branching point in the middle of the peptide chain

tected tetradepsipeptide, the heterocycle was introduced at and rarely at the C or N terminus. Usually, Ser is intro-

duced with its free hydroxy group, with careful selection of

mild coupling reagents such as DIPCDI/HOBt to prevent

acylation of the hydroxy functionality. However, it is some-

times convenient to have a protected Ser: when it is used as

a first amino acid, for instance, or if is coupled to a NMe-

amino acid. These couplings require stronger conditions,

and use of the Trt group has been described in these cases.

We also explored the use of the tert-butyldimethylsilyl

(TBDMS) group as a convenient hydroxy moiety protecting

group compatible with CTC resin. Although TBDMS is

widely used in organic and sugar chemistry,^[33]its use in Ser

or Thr protection has not been addressed.^[34]TBDMS can

be incorporated easily and can be removed on solid phase

by treatment with commercial TBAF solution. These condi-

tions appear to be compatible both with the peptidyl resin

bond and with the other protecting groups, such as Fmoc,

Boc, Alloc and pNZ. Boc--Ser(TBDMS)-OH was pre-

pared in solution from commercial Boc--Ser-OH and

TBDMS-Cl in the presence of a base (imidazole or DMAP,

64% yield). The amino acid Boc--Ser(TBDMS)-OH

was used in the tetradepsipeptide synthesis {[Alloc-Gly-

NMe-Cys(Acm)-NMe-Cys(Me)&][Boc--Ser(&)-O-CTC]}

and {[Alloc-Gly-NMe-Leu-NMe-Leu&][Boc--Ser(&)-O-

CTC]}, thereby showing the convenience of the TBDMS

hydroxy protecting group in SPPS.

d)

D

-Ser at the C Terminus in Fragment Coupling

When fragment coupling was performed with -Ser at

the C terminus (Scheme 13), three degradation products

were recovered instead of the octapeptide, thereby showing

the instability of the linear peptide chain. The byproducts

obtained – Fmoc--Ser-OH, the pentapeptide {[Fmoc--

Ser(OH)-Gly-Cys(Acm)-Cys(Me)-&][Fmoc--Ser(&)]} and

the tripeptide pNZ-Gly-Cys(Acm)-Cys(Me)-OH – were a

consequence of the breakage of the ester bonds. The second

Figure 6. HPLC profiles of a) quinaldic-, b) pNZ-, and c) 3-hy-

droxyquinaldic-protected tetradepsipeptides.

2966

www.eurjoc.org

Eur. J. Org. Chem. 2009, 2957–2974

Synthesis of Complex N-Methylated Depsipeptides

Scheme 13. Byproducts obtained in the fragment coupling starting at -Ser.

ester in particular was very labile, because no product con-

taining this bond was recovered. Because the first ester

bond is constructed at a very early stage, -Ser is not a

convenient option for the C terminus in cyclic depsipeptide

synthesis because the basic media and long reaction times

used in fragment couplings are incompatible with the sta-

bility of the linear chain.

2.3. Partial Oxidation of Cys(Me) – Cleavage

Optimization

Partial oxidation of Cys(Me) was observed during the

cleavage from the resin. Several cleavage cocktails were as-

sayed with the goal of minimizing the amounts both of oxi-

dized product and also of the Wang-linker derived product

(Figure 7). Addition of H₂O to the cleavage cocktail im-

proved the purity significantly. Cleavage of the tetradepsi-

peptide with TFA/CH₂Cl₂(1:4) thus resulted in a purity of

24%, increasing to 37% by cleavage with TFA/H₂O/CH₂Cl₂

(20:5:75), and to 77% with TFA/H₂O/CH₂Cl₂(2:1:7). The

last cleavage protocol was therefore applied in subsequent

syntheses on Wang resin.

3. Intermolecular Dimerization as a Peptide Chain

Elongation Step

So far, although we have identified several features to

improve the synthetic process, there is still no final strategy

that allows the preparation of oxathiocoraline. On the basis

of the purities of the obtained tetradepsipeptides, the use of

the strategy starting at Gly is the most promising. In this

approach, however, as well as in the rest of the strategies,

the main problem that remains to be solved is DKP forma-

tion when the peptide chain is elongated. As mentioned ear-

Figure 7. HPLC profiles for cleavage with a) TFA/CH₂Cl₂(1:4),

b) TFA/H₂O/CH₂Cl₂(20:5:75), and c) TFA/H₂O/CH₂Cl₂(2:1:7).

Eur. J. Org. Chem. 2009, 2957–2974

www.eurjoc.org

2967

N. Bayó-Puxan, J. Tulla-Puche, F. Albericio

FULL PAPER

lier, stepwise elongation entails a greater risk in terms of the effect of the protecting group in the dimerization step.

DKP formation because the second ester is positioned fur- In all cases, the formation of the dimer was achieved

ther away from the solid support and so is more flexible. In through two treatments with I₂(5 equiv., 10 min) in DMF,

the case of 4+4 fragment coupling strategies, approaches as shown in Scheme 14 and Figure 9.

that prevent ester bond formation during fragment coupling

either have associated DKP formation because of the pref-

erence for an intramolecular reaction (i.e., when starting at

Gly) or result in an unstable linear peptide (i.e., when start-

ing at -Ser). Since the use of distinct removal reagents for

protecting groups in stepwise or 4+4 fragment coupling

failed to prevent DKP formation, the only solution was to

restrict the mobility of the peptide chain by making the

NMe of Cys(Acm) unable to reach the carboxyl of NMe-

Cys(Me). This was achieved by the formation of the in-

terchain disulfide bridge. This dimerization through disul-

fide bond formation provided more rigidity and therefore

conferred more stability on the peptide chain, thereby pre-

venting DKP formation (Figure 8). Several optimizations

were performed to explore the potential of this approach

and to achieve high purities and yields in the dimer forma-

tion as the step prior to the final double cyclization.

Figure 9. HPLC profiles of a) Fmoc-protected tetradepsipeptide,

and b) tetradepsipeptide dimer.

We also evaluated dimer formation by the protected and

unprotected tetrapeptide when the peptide chain was

Figure 8. The intermolecular disulfide bond suppresses DKP for-

started with Gly. Scheme 15 shows that the first attempts

mation by restricting the mobility of the peptide chain.

were carried out with pNZ protection, Boc protection or

use of N-unprotected Cys(Acm). For these purposes, one

portion of the pNZ-protected resin-bound tetrapeptide was

3.1. Protected versus Unprotected Tetrapeptide

When the tetradepsipeptide chain was started at treated with SnCl₂to cleave the pNZ group, and the resins

Cys(Me), several options for the coupling of the last amino were treated with I₂(5 equiv.) in order to form the disulfide

acid were tested, such as the introduction of unprotected bridge. After two 10 min treatments, the degree of conver-

Boc--Ser-OH and also of Fmoc--Ser(Trt)-OH, followed sion to the protected dimer was only 17%, whereas full con-

by the introduction of the quinaldic unit. We then examined version to the unprotected dimer had been achieved. In

Scheme 14. Dimerization by disulfide bridge formation.

2968

www.eurjoc.org

Eur. J. Org. Chem. 2009, 2957–2974

Synthesis of Complex N-Methylated Depsipeptides

Scheme 15. Dimerization conversions with protected and unprotected tetrapeptides.

contrast, when the Boc-protected tetradepsipeptide was

used, conversion was complete, thereby suggesting that the

electronic behaviour of the pNZ group impedes the progress

of the reaction. In view of the results, the best and safest

option is to use the Boc-protected tetradepsipeptide. This

approach presents three advantages: Boc--Ser-OH is com-

mercially available, the dimer is quantitatively formed, and

the cleavage from the resin simultaneously removes the Boc

Scheme 16. Bis-lactonization from the dimer precursor.

groups, thereby providing the unprotected dimer of tetra-

peptides ready for the final cyclization step.

Lactonization was tested with two coupling systems com-

monly used for ester formation: DIPCDI/DMAP and

MSNT/DIEA. For the Fmoc- and Boc-protected dimers,

3.2. Double Lactonization – Reactivity in Ester Formation

as a Function of the Amino Protecting Group of D-Ser

As described above, when the tetradepsipeptide was the formation of bicyclic depsipeptide was observed with

started at Cys(Me), the -Ser residue was introduced as a the DIPCDI/DMAP system, although the rate of cycliza-

fourth amino acid in several protected forms, and dimeriza- tion was slow and complete cyclization was not achieved.

tion was carried out. The next step consisted of the double Although MSNT/DIEA^[35]is a more potent coupling sys-

lactonization of the dimer to afford the bicyclic skeleton tem for esterification than DIPCDI/DMAP, cyclization was

(Scheme 16). The three protected dimers (Fmoc, Boc and not detected. One possible explanation for this observation

quinaldic acid) were tested to examine the influences of: is that the excess of base used in this coupling breaks the

a) carbamate versus amide functionality, and b) distinct ester functionality. This notion is supported by the observa-

protecting groups.

tion that attempts to remove the Fmoc group from the pro-

Figure 10. Reactivity in the formation of the ester bond depending on the group present at the amino terminus of -Ser.

Eur. J. Org. Chem. 2009, 2957–2974

www.eurjoc.org

2969

N. Bayó-Puxan, J. Tulla-Puche, F. Albericio

FULL PAPER

tected bicyclic depsipeptide with piperidine or diethylamine formation and also the difficulty in removing the protecting

resulted in loss of the compound as a result of degradation. group at the -Ser side-chain (Figure 10).

When quinaldic acid protection was present, no cyclization

Given the problems of low reactivity encountered in the

to form the bicyclic depsipeptide was observed in any cou- double lactonization, our efforts focused on strategies with

pling system.

lactamization as the final step.

Thus, from these results and from the behaviour ob-

served during the elongation of the tetrapeptidyl chain with

use of the different protection schemes, we conclude that

the reactivity of the ester bond is greatly influenced by the

3.3. Lactamization – Obtaining Demethylated

Oxathiocoraline Analogues

When the peptide chain was started at Gly on Wang

nature of the group (amide or carbamate) present at the resin, Boc-Cys(Acm)-OH was introduced as the fourth

amino terminus of the Ser residue. With regard to the car- amino acid. As described above, the two tetradepsipeptides

bamate-type protecting groups, the electron-withdrawing (one bearing the quinaldic acid and the other the natural 3-

capacity of the group determines the reactivity in the ester hydroxyquinaldic acid) were obtained in good purities (Fig-

Scheme 17. Final steps in the synthesis of demethylated analogues of oxathiocoraline.

Figure 11. Structures, UV spectra and HPLC analysis of the final demethylated analogues of oxathiocoraline.

2970

www.eurjoc.org

Eur. J. Org. Chem. 2009, 2957–2974

Synthesis of Complex N-Methylated Depsipeptides

Scheme 18. Disulfide reduction by addition of DTT.

ure 6). The two tetradepsipeptides were subsequently di-

In the course of the studies directed towards the synthesis

merized, cleaved and subjected to lactamization by treat- of oxathiocoraline, we propose these guidelines for the as-

ment with PyBOP and HOAt in DMF at pH 8 (DIEA) sembly of other cyclodepsipeptides:

(Scheme 17). The lactamization proceeded smoothly and

was complete in 1 h. After workup and purification by

semipreparative HPLC, two analogues of oxathiocoraline,

{[3-HQA--Ser(&¹)-Gly-Cys(&²)-Cys(Me)&³] [3-HQA--

Ser(&³)-Gly-Cys(&²)-Cys(Me)&¹]} and {[QNA--Ser(&¹)-

a) If possible, to simplify the synthetic approach by tak-

ing the symmetry of the molecule into account.

b) To minimize the risk of DKP formation by: i) avoiding

Cys(Me) as the C terminus because it is very prone to give

DKPs as a result of the acidity of its β-proton, ii) using

Alloc and pNZ groups instead of the Fmoc group,^[19,20]

iii) using CTC resin, which because of its steric hindrance

gives better results than Wang resin,^[15,16]and iv) restricting

the flexibility of the peptide chain by, for example, using

the dimerization approach.^[4]

c) To choose the chemistry carefully after the ester bond

has been formed. The Fmoc group should be avoided and

the pNZ group can be used only at certain positions; it is

not compatible, for instance, as a protecting group for the

amino acid to be coupled as an ester because its removal

results in hydrolysis of the ester bond. Neither is pNZ com-

patible as a protecting group for the first amino acid an-

chored to CTC resin.

Gly-Cys(&²)-Cys(Me)&³]

[QNA--Ser(&³)-Gly-Cys(&²)-

Cys(Me)&¹]}, were obtained (Figure 11).

3.4. Reduction of the Disulfide Bridge with DTT

To rule out the possibility of two cyclotetradepsipeptides

linked by a disulfide bridge (Scheme 18, A), the quinaldic

acid cyclodepsipeptide was subjected to disulfide reduction

by exposure to DTT in a denaturating buffer (6  Gdn·HCl,

1 m EDTA, 0.1  Tris·HCl at pH 8.7) under N₂

(Scheme 18). Under these conditions, only a compound

with two units of mass more than the bicyclic depsipeptide

was observed, which corresponded to the reduced octacy-

clodepsipeptide (Scheme 18, B).

d) Tandem deprotection/coupling with Alloc- and Fmoc-

AA-Fs are not compatible with CTC resins because partial

detachment of the peptide chain occurs.

Conclusion and Final Strategy

e) NMe-Cys(Me) is not a good choice as first amino acid

when using a Wang resin, which requires TFA (50%) for

final cleavage, because these conditions favour de-N-methyl-

ation.

f) The protecting group for the α-amino functionality is

relevant to the reactivity of the β-position because it may

substantially enhance the acidity of the α-proton. The rate

of esterification/lactonization of the Ser hydroxy group, the

yields of the β-elimination reactions of both Cys and even

Ser, and the oxidation of the thiol group are all dependent

on the α-amino-protecting group. Acyl is the poorest

choice.

A number of factors must be considered when at-

tempting the synthesis of a sophisticated depsipeptide. The

main limitations arise from the ester functionality, which, in

combination with the presence of consecutive NMe-amino

acids, numerous Cys residues and a disulfide bridge, re-

quires careful examination and optimization of each step.

DKP formation and approaches to overcome this process

were explored in several situations. The main trends in the

instability and reactivity of the ester bonds were studied, as

well as the side-reactions associated with Cys. The new con-

cept of restricting mobility to protect the ester bond against

DKP formation was used first to obtain demethylated ana-

logues of oxathiocoraline, and later for the synthesis of oxa-

thiocoraline (Scheme 19).

g) Ser is not a suitable choice as C terminus because it

makes the β-ester highly labile. This negative effect is fur-

Eur. J. Org. Chem. 2009, 2957–2974

www.eurjoc.org

2971

N. Bayó-Puxan, J. Tulla-Puche, F. Albericio

FULL PAPER

Scheme 19. Synthesis of oxathiocoraline by a fragment dimerization approach.

Loading of the First Amino Acid onto Wang Resin: The resin was

first washed with DMF (3ϫ1 min) and CH₂Cl₂(3ϫ1 min). Next,

the Fmoc-AA-OH (8 equiv.) was dissolved in CH₂Cl₂/DMF (9:1)

and DIPCDI (4 equiv.) was added. The mixture was then added to

the resin, DMAP (1 equiv.) in DMF (0.2 mL) was finally added,

and the mixture was allowed to react for 14 h at 25 °C.

ther increased when the protecting group of the N-terminal

Ser is an acyl moiety.

h) The TBDMS group is an alternative to the usual Trt

or tBu groups that can be used in conjunction with CTC

resin and Fmoc, pNZ and Boc protecting groups.^[34]

i) Gly is not a suitable choice as a residue next to the C

terminus in a peptide fragment to be coupled because its

poor hindrance favours the formation of intramolecular

DKPs. This side reaction will compete with the condensa-

tion, thereby resulting in a low reaction yield.

j) To couple a NMe-AA but at the same to minimize pos-

sible side-reactions, a reactive but neutral coupling (HATU

with a deficit of base) gives the best yields.

k) Optimization of the cleavage cocktail should be car-

ried out to reduce oxidation byproducts.

Peptide Elongation: Protected amino acids (5 equiv.) were coupled

either with HATU/HOAt or HBTU/HOBt (5 equiv. each) and

DIEA (10 equiv.) in DMF for 1 h. In cases in which a neutral cou-

pling was necessary, the coupling system DIPCDI/HOAt (5 equiv.

each) in DMF was used.

Fmoc Group Removal: The peptide-resin was treated with piper-

idine/DMF (1:4, 2ϫ1 min, 2ϫ5 min) and washed with DMF

(3ϫ1 min) and CH₂Cl₂(3ϫ1 min).

Alloc Group Removal: The peptide-resin was treated under Ar with

PhSiH₃(10 equiv.) and [Pd(PPh₃)₄] (0.1 equiv.) dissolved in CH₂Cl₂

for 15 min, and was then washed with CH₂Cl₂(3ϫ1 min). The

process was repeated three times.

The points outlined above have been observed in other

depsipeptides and form the basis of the design of synthetic

approaches in our laboratory.

pNZ Group Removal: The peptide resin was treated with SnCl₂

(6 ) and HCl (1.6 m) in DMF for 30 min. The resin was then

washed with DMF (3ϫ1 min) and CH₂Cl₂(3ϫ1 min), and the

process was repeated once.

Experimental Section

Loading of the First Amino Acid onto CTC Resin: The resin was

first washed with DMF (5ϫ1 min) and CH₂Cl₂(3ϫ1 min) and a

solution of Fmoc-AA-OH (1 equiv.) and DIEA (6.7 equiv.) in

CH₂Cl₂was added. After 10 min, more DIEA (3.3 equiv.) was

added and the mixture was stirred for 50 min at room temperature.

The reaction was quenched by the addition of MeOH (0.4 mL per

g resin) and the mixture was stirred for a further 10 min.

Trt Group Removal: The peptide resin was treated with TFA/TIS/

CH₂Cl₂(2:2.5:95.5, 5ϫ1 min). The resin was then washed with

DMF (3ϫ1 min) and CH₂Cl₂(3ϫ1 min).

TBDMS Group Removal: The peptide was treated with a solution

of TBAF in THF (1 , 3ϫ10 min). The resin was then washed

with THF (3ϫ1 min), DMF (3ϫ1 min) and CH₂Cl₂(3ϫ1 min).

2972

www.eurjoc.org

Eur. J. Org. Chem. 2009, 2957–2974

Synthesis of Complex N-Methylated Depsipeptides

[4]

[5]

J. Tulla-Puche, N. Bayó-Puxan, J. A. Moreno, A. M. Fran-

cesch, C. Cuevas, M. Álvarez, F. Albericio, J. Am. Chem. Soc.

2007, 129, 5322–5323.

N. Bayó-Puxan, A. Fernández, J. Tulla-Puche, E. Riego, C.

Cuevas, M. Álvarez, F. Albericio, Chem. Eur. J. 2006, 12, 9001–

9009.

N. Bayó-Puxan, A. Fernández, J. Tulla-Puche, E. Riego, M.

Álvarez, F. Albericio, Int. J. Pept. Res. Ther. 2007, 13, 295–306.

F. Romero, F. Espliego, J. Pérez Baz, T. García de Quesada, D.

Gravalos, F. De la Calle, J. L. Fernández-Puentes, J. Antibiot.

1997, 50, 734–737.

J. Pérez Baz, L. M. Cañedo, J. L. Fernández Puentes, M. V.

Silva Elipe, J. Antibiot. 1997, 50, 738–741.

E. Atherton, P. M. Hardy, D. E. Harris, B. H. Matthews in Pep-

tides 1990, Proceedings of the 21st European Peptide Sympo-

sium (Eds.: E. Giralt, D. Andreu), ESCOM Science Publishers

B. V., Leiden, The Netherlands, 1991, pp. 243–244.

Y. Fujiwara, K. Akaji, Y. Kiso, Chem. Pharm. Bull. 1994, 42,

724–726.

J. Lukszo, D. Patterson, F. Albericio, S. A. Kates, Lett. Pept.

Sci. 1996, 3, 157–166.

G. Barany, Y. Han, B. Hargittai, R.-Q. Liu, J. T. Varkey, Bio-

polymers 2003, 71, 652–666.

P. Lloyd-Williams, F. Albericio, E. Giralt, Chemical Approaches

to the Synthesis of Peptides and Proteins, CRC, Boca Raton

(FL, USA), 1997.

K. Barlos, O. Chatzi, D. Gatos, G. Stavrospoulos, Int. J. Pept.

Protein Res. 1991, 37, 513–520.

P. Rovero, S. Vigano, S. Pegoraro, L. Quartara, Lett. Pept. Sci.

1996, 2, 319–323.

C. Chiva, M. Vilaseca, E. Giralt, F. Albericio, J. Pept. Sci.

1999, 5, 131–140.

M. Gairi, P. Lloyd-Williams, F. Albericio, E. Giralt, Tetrahe-

dron Lett. 1990, 31, 7363–7366.

J. Alsina, E. Giralt, F. Albericio, Tetrahedron Lett. 1996, 37,

4195–4198.

N. Thieriet, J. Alsina, E. Giralt, F. Guibé, F. Albericio, Tetrahe-

dron Lett. 1997, 38, 7275–7278.

A. Isidro-Llobet, J. Guasch-Camell, M. Alvarez, F. Albericio,

Eur. J. Org. Chem. 2005, 14, 3031–3039.

Solid-Phase N-Methylation: After removal of the corresponding

amino-protecting group, a solution of o-NBS-Cl (4 equiv.) and

DIEA (10 equiv.) in CH₂Cl₂was added to the resin and the mixture

was stirred for 90 min. After filtration and washings with CH₂Cl₂

(3ϫ1 min), DMF (3ϫ1 min), CH₂Cl₂(3ϫ1 min) and THF

(3ϫ1 min), a solution of PPh₃(5 equiv.) and MeOH (10 equiv.) in

THF and a solution of DIAD (5 equiv.) in THF were mixed and

added to the peptide resin. After the resin had been stirred for 1 h,

it was washed with THF (3ϫ1 min), CH₂Cl₂(3ϫ1 min) and DMF

(3ϫ1 min). After treatments (2ϫ15 min) with DBU (5 equiv.) and

2-mercaptoethanol (10 equiv.) in DMF, the resin was washed with

DMF (3ϫ1 min), CH₂Cl₂(3ϫ1 min) and DMF (3ϫ1 min).

[6]

[7]

[8]

[9]

Tandem Deprotection/Coupling Reaction: The peptide-resin was

treated with PhSiH₃(12 equiv.), [Pd(PPh₃)₄] (0.2 equiv.) and pre-

viously prepared Fmoc-AA-F (7 equiv.) in CH₂Cl₂for 15 min, and

more Fmoc-AA-F (7 equiv.) was then added. The resin was finally

washed with CH₂Cl₂(3ϫ1 min).

[10]

[11]

[12]

[13]

Fast Neutral Removal/Coupling Step: To remove the Alloc group

rapidly, the peptide-resin was treated under Ar for 10 min with

PhSiH₃(10 equiv.) and [Pd(PPh₃)₄] (0.1 equiv.) dissolved in

CH₂Cl₂, and was then washed with CH₂Cl₂(3ϫ1 min). The pro-

cess was repeated once, and two consecutive fast couplings of the

Alloc-AA-OH (5 equiv.) with HATU/HOAt (5 equiv. each) and

DIEA (4.8 equiv.) in DMF (35 min) were then carried out.

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21]

Solid-Phase Dimerization: Formation of the intermolecular disul-

fide bridge was achieved by treatment of the resin-bound tetrapep-

tide with I₂(5 equiv., 0.06 ) in DMF (2ϫ10 min). The resin was

then repeatedly washed with CH₂Cl₂(10ϫ1 min), DMF

(10ϫ1 min) and CH₂Cl₂(10ϫ1 min).

Reduction of the Disulfide Bridge: The peptide (1.9 µmol) was dis-

solved under N₂in a denaturating buffer (6  Gdn·HCl, 1 m

EDTA, 0.1  Tris·HCl at pH 8.7, 1 mL). A dithiothreitol (DTT)

solution (0.38 , 0.5 mL) was then added, and the reaction mixture

was stirred for 45 min at 25 °C.

Cleavage from CTC Resin: The peptide was cleaved from the resin

by treatment with a TFA/CH₂Cl₂solution (2:98, 5ϫ1 min) and the

filtrates were collected in the presence of H₂O (60 mL per g of

resin), dried and lyophilized.

For exploratory studies through this work, non-NMe-Cys resi-

dues and/or N-methylated commercial amino acids such as

NMe-Leu were used instead of the NMe-Cys, because of avail-

ability problems.

A. Madder, N. Farcy, N. G. C. Hosten, H. De Muynck, P. J.

De Clercq, J. Barry, A. P. Davis, Eur. J. Org. Chem. 1999, 2787–

2791.

M. J. O. Anteunis, C. Van der Auwera, Int. J. Pept. Protein Res.

1988, 31, 301–310.

M. Teixidó, F. Albericio, E. Giralt, J. Pept. Res. 2005, 65, 153–

166.

Cleavage from Wang Resin: The peptide was cleaved from the resin

by treatment with TFA/H₂O/CH₂Cl₂(3:1:6, 5 mL, 2ϫ30 min), the

solvents were removed from the mixture (N₂), and the product was

lyophilized.

[22]

[23]

[24]

[25]

[26]

[27]

Supporting Information (see also the footnote on the first page of

this article): Detailed experimental procedures and characterization

material for Boc--Ser(TBDMS)-OH and demethylated analogues

of oxathiocoraline.

S. C. Miller, T. S. Scanlan, J. Am. Chem. Soc. 1998, 120, 2690–

2691.

For nomenclature see; J. Spengler, J. C. Jiménez, K. Burger, E.

Giralt, F. Albericio, J. Pept. Res. 2005, 65, 550–555.

Phosphonium salts and DIEA are used as coupling reagents

instead of uronium salts to avoid guanylidation side-products

that may occur during long couplings (F. Albericio, J. M. Bof-

ill, A. El-Faham, S. A. Kates, J. Org. Chem. 1998, 63, 9678–

9683).

M. T. Hamann, P. J. Scheuder, J. Am. Chem. Soc. 1993, 115,

5825–5826.

N. D. Fulton, K. Bollenbacher, G. E. Templeton, Phytopathol-

ogy 1965, 55, 49–51.

Acknowledgments

This study was partially supported by PharmaMar (Madrid), the

Comisión Interministerial de Ciencia y Tecnología (CTQ2006-

03794/BQU), the Carlos III Health Institute (CIBER, nanomedic-

ine), the Institute for Research in Biomedicine, and the Barcelona

Science Park. J. T.-P. is a Juan de la Cierva fellow (MICINN).

[28]

[29]

[30]

[31]

[1] F. Sarabia, S. Shammaa, A. Sanchez Ruiz, L. Martin Ortiz,

F. J. Lopez Herrera, Curr. Med. Chem. 2004, 11, 1309–1332.

[2] Y. Hamada, T. Shioiri, Chem. Rev. 2005, 105, 4441–4482.

[3] J. Fotie, R. E. Morgan, Mini-Rev. Med. Chem. 2008, 8, 1088–

1094.

F. Cavelier, J. Verducci, F. Andre, F. Haraux, C. Sigalat, M.

Traris, A. Vey, Pestic. Sci. 1998, 52, 81–89.

A. Isogai, N. Fukuchi, S. Yamashita, K. Suyama, A. Suzuki,

Tetrahedron Lett. 1990, 31, 695–698.

Eur. J. Org. Chem. 2009, 2957–2974

www.eurjoc.org

2973

N. Bayó-Puxan, J. Tulla-Puche, F. Albericio

FULL PAPER

[32] J. C. Jiménez, N. Bayó, B. Chavarría, A. López-Macia, M.

Royo, E. Nicolás, E. Giralt, F. Albericio, Lett. Pept. Sci. 2002,

9, 135–141.

[33] P. G. M. Wuts, T. W. Greene, Greene’s Protective Groups in Or-

ganic Synthesis, 4^thed., John Wiley & Sons, Inc., Hoboken (NJ,

USA), 2007.

[34] TBS has been used for protection of Ser: W. R. Ewing, M. M.

Joullié, Heterocycles 1988, 27, 2843–2850.

[35] B. Blankmeyer-Menge, M. Nimtz, R. Frank, Tetrahedron Lett.

1990, 31, 1701–1704.

Received: March 11, 2009

Published Online: May 12, 2009

2974

www.eurjoc.org

Eur. J. Org. Chem. 2009, 2957–2974

Article Doi

Oxathiocoraline: Lessons to be learned from the synthesis of complex N-methylated depsipeptides

DOI: 10.1002/ejoc.200900259

Source and publish data:

Authors:

Article abstract of DOI:10.1002/ejoc.200900259

Full text of DOI:10.1002/ejoc.200900259

Products guided by the article

R&D Labs maybe for 1176846-54-0

Relevant to this article

Hot Product