2-methoxy-4-methylsulfinylbenzyl: A backbone amide safety-catch protecting group for the synthesis and purification of difficult peptide sequences

Article abstract of DOI:10.1002/chem.201403668

The use of 2-methoxy-4-methylsulfinylbenzyl (Mmsb) as a new backbone amide-protecting group that acts as a safety-catch structure is proposed. Mmsb, which is stable during the elongation of the sequence and trifluoroacetic acid-mediated cleavage from the resin, improves the synthetic process as well as the properties of the quasi-unprotected peptide. Mmsb offers the possibility of purifying and characterizing complex peptide sequences, and renders the target peptide after NH₄I/TFA treatment and subsequent ether precipitation to remove the cleaved Mmsb moiety. First, the "difficult peptide" sequence H-(Ala)₁₀-NH₂ was selected as a model to optimize the new protecting group strategy. Second, the complex, bioactive Ac-(RADA)₄-NH₂ sequence was chosen to validate this methodology. The improvements in solid-phase peptide synthesis combined with the enhanced solubility of the quasi-unprotected peptides, as compared with standard sequences, made it possible to obtain purified Ac-(RADA)₄-NH₂. To extend the scope of the approach, the challenging Aβ(1-42) peptide was synthesized and purified in a similar manner. The proposed Mmsb strategy opens up the possibility of synthesizing other challenging small proteins.

Full text of DOI:10.1002/chem.201403668

DOI: 10.1002/chem.201403668
Full Paper
&
Solid-Phase Synthesis
2-Methoxy-4-methylsulfinylbenzyl: A Backbone Amide Safety-
Catch Protecting Group for the Synthesis and Purification of
Difficult Peptide Sequences
Marta Paradꢀs-Bas,[a, b] Judit Tulla-Puche,*[a, b] and Fernando Albericio*[a, b, c, d]
Abstract: The use of 2-methoxy-4-methylsulfinylbenzyl
(Mmsb) as a new backbone amide-protecting group that
acts as a safety-catch structure is proposed. Mmsb, which is
stable during the elongation of the sequence and trifluoro-
acetic acid-mediated cleavage from the resin, improves the
synthetic process as well as the properties of the quasi-un-
protected peptide. Mmsb offers the possibility of purifying
and characterizing complex peptide sequences, and renders
the target peptide after NH4I/TFA treatment and subsequent
ether precipitation to remove the cleaved Mmsb moiety.
First, the “difficult peptide” sequence H-(Ala)10-NH2 was se-
lected as a model to optimize the new protecting group
strategy. Second, the complex, bioactive Ac-(RADA)4-NH2 se-
quence was chosen to validate this methodology. The im-
provements in solid-phase peptide synthesis combined with
the enhanced solubility of the quasi-unprotected peptides,
as compared with standard sequences, made it possible to
obtain purified Ac-(RADA)4-NH2. To extend the scope of the
approach, the challenging Ab(1-42) peptide was synthesized
and purified in a similar manner. The proposed Mmsb
strategy opens up the possibility of synthesizing other
challenging small proteins.
Introduction
These kinds of peptides are characterized by incomplete re-
moval of the amino protecting group and poor incorporation
of the protected amino acids at some stage of the synthesis,
which prevents peptide elongation even after repeating
reaction treatment. It is accepted that these difficulties are due
to the formation of peptide secondary structure, such as
b-sheets.[12,13] During the synthesis, these intra- or inter-
molecular interactions, which are usually sequence-dependent,
spontaneously form aggregations. These assemblies are
attributed to the formation of the backbone amide hydrogen
bond with carbonyl groups of peptide chains also bound to
the resin.
Solid-phase peptide synthesis (SPPS) is nowadays the most
common strategy used to synthesize biopolymers in high yield
and purity. This approach has been made possible thanks to
great efforts of the scientific community to develop a wide
range of functionalized resins,[1] coupling reagents,[2–5] and or-
thogonal protecting groups.[6,7] After extensive fine-tuning
over the last 50 years, one would assume that chemists would
have all the tools necessary to prepare a short sequence of
any nature. However, this is not the case. The so-called “diffi-
cult sequences”, which are short sequences that are apparently
unreachable targets,[8–11] continue to pose a challenge.
Self-interactions require the presence of the NH of the pep-
tide bond, therefore, Pro-containing peptide sequences are rel-
atively free of this problem. In this regard, Mutter described
a revolutionary concept in protecting group strategy: the
pseudoproline. This concept converts Ser, Thr, and Cys into
their five-membered ring dimethyl derivatives, thi/ox-
azolidines, which, during the final global deprotection, render
the native residues.[14,15] Following the same idea of masking
the presence of the NH moiety, Kiso, Carpino, and Beyermann
independently proposed the synthesis of Ser and Thr depsi-
peptides, which are more manageable (soluble building
blocks) and can be converted into the corresponding peptides
in basic medium.[16–19]
[a] M. Paradꢀs-Bas, Dr. J. Tulla-Puche, Prof. F. Albericio
Institute for Research in Biomedicine (IRB Barcelona)
Baldiri Reixac 10, Barcelona, 08028 (Spain)
Fax: (+34)93-4037126
E-mail: judit.tulla@irbbarcelona.org
albericio@irbbarcelona.org
[b] M. Paradꢀs-Bas, Dr. J. Tulla-Puche, Prof. F. Albericio
CIBER-BBN, Networking Centre on Bioengineering
Biomaterials and Nanomedicine, Barcelona Science Park
Baldiri Reixac 10, Barcelona, 08028 (Spain)
[c] Prof. F. Albericio
Department of Organic Chemistry, University of Barcelona
Martꢀ i Franquꢁs 1-11, Barcelona, 08028 (Spain)
Parallel to the work of Mutter, Sheppard and co-workers put
forward the use of an amide-protecting group, N-(2-hydroxy-4-
methoxybenzyl) (Hmb).[20,21] However, during activation, the
building block Na-Fmoc-Na-(Hmb)amino acid forms a 4,5-di-
hydro-8-methoxy-1,4-benzoxapin-2(3H)-one, which shows poor
reactivity.[22] Thus, these residues should be incorporated in the
[d] Prof. F. Albericio
School of Chemistry & Physics, University of Kwazulu-Natal
University Road, Westville, X54001, Durban, 4001 (South Africa)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201403668.
Chem. Eur. J. 2014, 20, 15031 – 15039
15031
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
form of O,N-bis(Fmoc) derivatives. Once these building blocks
have been introduced, and after removal of both Fmoc
groups, the coupling of the incoming amino acid is facilitated
through a seven-membered ring O!N transacylation mecha-
nism. In addition, the same authors proposed the use of a so-
called “reversible backbone amide” protecting group, N-(3-
methylsulfinyl-4-methoxy-6-hydroxybenzyl) (SiMB). In addition
to the free hydroxyl group, SiMB contains an electron-with-
drawing sulfoxide substituent para to the OH, with the unique
purpose of enhancing the acylation step.[23] Global deprotec-
tion is carried out through a reductive acidolysis to convert
the sulfoxide into the thioether and then allow the amide pro-
tecting group to be cleaved by trifluoroacetic acid (TFA). Simi-
larly, the so-called “reversible” backbone protecting groups
based on the nitro group, which is cleavable under photolysis,
were proposed by Kent and co-workers.[24]
Figure 1. The concept of the new backbone amide-protecting group
(Mmsb).
sophisticated linkers based on the same principle have also
been used for SPPS.[36]
Here, we used the Mmsb group to prepare three “difficult
sequences”. After TFA global deprotection, the Mmsb-contain-
ing peptides were characterized and purified, when necessary,
before performing a reductive TFA treatment, which reduces
the sulfoxide to thioether and concomitantly removes it in
a one-pot reaction.
During recent years, our group has focused on the synthesis
of these “difficult sequences”. In this regard, we have used di-
methoxybenzyl (Dmb) and developed 1-methyl-3-indolylmeth-
yl (MIM) and 3,4-ethylenedioxy-2-thenyl (EDOTn)[25] to reduce
intra- and inter-chain interaction, as well as prevent side reac-
tions associated with the presence of the NH group, such as
aspartimides[26–28] and internal diketopiperazine formation.[29,30]
The design of the last two protecting groups mentioned has
shown some advantages over benzyl derivatives. However, we
have focused our efforts on a more global solution for the syn-
thesis and on the characterization and purification of complex
sequences that is not limited to Ser/Thr/Cys but valid for
a large number of residues. Here we describe a “safety-catch”
backbone amide-protecting group[31] that is compatible not
only with the 9-fluorenylmethoxycarbonyl (Fmoc) strategy, but
also with TFA treatment, with the latter being used to detach
the peptide from the resin and remove the classical side-chain
protecting groups. The synthesis of a quasi-protected peptide,
which still bears the safety-catch protecting group, facilitates
the characterization and purification of the peptide, and allows
the target free-peptide to be rendered after a final and clean
chemical treatment.
First, the H-(Ala)10-NH2 sequence was selected as a model to
study and optimize the new synthetic methodology proposed
herein. Second, two interesting bioactive peptides, Ac-(RADA)4-
NH2 (RADA-16) and Ab(1-42), both with a high tendency to
self-assemble, were synthesized for the validation of the Mmsb
strategy, established in the first model. The first sequence is
a challenging peptide for SPPS because of its hydrophobicity
and its high tendency to form secondary structures that jeop-
ardize the deprotection of the amino function and also acyla-
tion.[37] This peptide has been widely used to assay novel syn-
thetic approaches.[38,39] The second peptide, RADA-16, is of in-
terest because of its biomedical applications,[40,41] but it is
again associated with a complex synthesis.[42] In this case, ag-
gregation is related to its high content of alternating positively
and negatively charged amino acids. Finally, we addressed the
synthesis of the molecule considered to be responsible for
amyloid formation in Alzheimer’s disease, the extensively
studied Ab(1-42).[43,44] The synthesis of this peptide is
challenging[18,45,46] because this molecule shows a strong
tendency to form aggregates, which directly affects the
elongation sequence and, at the same time, its solubility, thus
complicating its purification.
The Fmoc SPPS strategy is based on the use of a temporary
protecting group that is removed by a base (piperidine) and
cleavage upon global deprotection (by TFA), therefore, we
chose not introduce any extra chemical reaction. We envisaged
that the pair alkylsulfoxide/thioether would serve our require-
ments. Thus, the electron-withdrawing sulfoxide group intro-
duces stability to a benzyl moiety, and it can be easily reduced
to the corresponding electron-donating thioether, which con-
fers acid lability to the same benzyl group. Specifically, 2-me-
thoxy-4-methylsulfinylbenzyl (Mmsb), which is stable to TFA,
was reduced to 2-methoxy-4-methylthiobenzyl (Mmtb), which
is labile to TFA, thus perfectly matching our requirements
(Figure 1).
Results and Discussion
To introduce the new Mmsb backbone amide-protecting
group into the peptide sequence, the corresponding Fmoc de-
rivative (Fmoc-N(Mmsb)-Ala-OH in our case) was prepared. The
synthesis of this building block was accomplished in five steps
(Scheme 1, see the Supporting Information for details).[47] Com-
mercially available 3-methoxythiophenol (1) was methylated in
practically quantitative yield with iodomethane (MeI) by careful
dropwise addition of triethylamine to prevent dialkylation.
Next, alkylated product 2 was formylated by reaction with
freshly prepared Vilsmeier reagent,[48] which rendered the two
possible isomers. 2-Methoxy-4-methylthiobenzaldehyde (3)
was isolated (48% yield) after purification by reverse-phase
chromatography on a C18 column. The reductive amination of
aldehyde 3 with the amine of the unprotected alanine was
reached in a one-step reaction with NaBH3CN in dioxane/H2O
(1:1).[25] N(Mmtb)-Ala-OH (4) was protected with the Fmoc
The Mmsb moiety has been used for the preparation of the
corresponding carbamate-based amine protection and inte-
grated as a linker for the preparation of peptide acids.[32] The
demethoxy (p-methylsulfinylbenzyl) version was also used as
a carboxyl protecting group[33] and as a linker,[34] and its
corresponding carbamate-based amine protection.[35] More
Chem. Eur. J. 2014, 20, 15031 – 15039
www.chemeurj.org
15032
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Ala-OH was introduced with only 1.5 equiv (AA/coupling re-
agent), and the best conditions for the key incorporation of
the incoming residue consisted of three couplings with Fmoc-
Ala-OH (10 equiv), DIPCDI (10 equiv), and OxymaPure
(10 equiv) at 458C, with MeCN/N,N-dimethylformamide (DMF)
(3:1) as solvent. Once again, we observed that MeCN is a suita-
ble solvent for incorporations onto hindered amino acids.[50]
Except for the coupling of the incoming Ala on the N-Mmsb
residue, no double couplings were performed. However, the
three peptide elongations were checked by the Kaiser test[51]
after each coupling. The tests revealed that synthesis of the
standard peptide took place with incomplete incorporations of
the last four Ala residues.
Scheme 1. Synthesis of Fmoc-N(Mmsb)-Ala-OH. O/N=overnight.
The two modified Mmsb-containing peptides were cleaved
from the resin by using TFA/CH2Cl2 (19:1), rendering both se-
quences with similar purity (Figures S3 and S4 in the Support-
ing Information), and the total stability of the Mmsb group
was confirmed by MS analysis (m/z calcd for C39H63N11O12S:
909.44; found: 932.60 [M+Na]+ (8) and 932.45 [M+Na]+ (9)).
Finally, we used MS analysis to study the reductive acidolysis
of the Mmsb-containing peptides caused by treatment with
NH4I/TFA (Figure 3 and Figure S12 in the Supporting Informa-
tion). It was found that the Mmsb group was completely re-
moved after 30 min in the case of 8 and after 2 h in the case
of 9. This deprotection was accomplished without the need for
group by using a slight excess of Fmoc-OSu in basic media to
afford Fmoc-N(Mmtb)-Ala-OH (5), which was easily purified by
reverse-phase chromatography on a C18 column to remove
the excess of unreacted alanine during the reductive amination
step. Finally, oxidation of
5 with H2O2 rendered Fmoc-
N(Mmsb)-Ala-OH (6) in excellent purity (98.0%) and without
further purification; this building block was used directly for
SPPS. The overall yield to convert 1 into 6 was 19%.
First, we tested our synthetic strategy on H-(Ala)10-NH2. As
a control, the standard peptide was also synthesized with only
the use of Fmoc-Ala-OH (7; Figure 2). Two modified H-(Ala)10-
NH2 sequences were synthesized with one Fmoc-N(Mmsb)-Ala-
OH at positions 6 (8) and 8 (9), respectively.
The three peptides were synthesized on Rink amide-poly-
styrene (PS) as a solid support and N,N’-diisopropylcarbodi-
imide (DIPCDI)/ethyl 2-cyano-2-(hydroxyimino)acetate (Oxyma-
Pure)[49] as coupling reagent. In this regard, Fmoc-N(Mmsb)-
scavengers (some articles report the use of Me2S)[23,32]
.
One proposed solution to prevent the main impurity detect-
ed in both modified Mmsb-containing peptide syntheses (9%
in 8 and 6% in 9) associated with the truncated peptide after
coupling of the N-Mmsb residue involved preparing the dipep-
tide (in this case Fmoc-Ala-N(Mmsb)-Ala-OH) in solution (see
the Supporting Information for details). This strategy was used
Figure 2. Structures of the synthesized peptides.
Chem. Eur. J. 2014, 20, 15031 – 15039
www.chemeurj.org
15033
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 4. MALDI-TOF MS and solubility assays of H-(Ala)10-NH2 crude pep-
tides synthesized: a) by using the common Ala, 7, and b) by using the modi-
fied Fmoc-N(Mmsb)-Ala-OH, 8. Concentration: 1 mgmLꢀ1 in H2O. Mass im-
purities assigned (red label), mass expected peptide (green label).
a b-sheet structure profile (minimum at 215 nm), characteristic
of self-assembling sequences, whereas the two modified pep-
tides (8 and 9) showed a random coil structure (minimum at
200 nm). These data suggested the involvement of the Mmsb
in the secondary structure of this peptide, possibly explaining
the synthetic improvements observed in modified H-(Ala)10-
NH2.
Encouraged by the results obtained in this peptide model
using the new method, we explored the application of Mmsb
to other “difficult peptides”.
Figure 3. Monitoring of the Mmsb group removal by MALDI-TOF MS spectra.
Starting material (Mmsb-protected H-(Ala)10-NH2) 8 is highlighted in blue,
whereas the reaction intermediate (Mmtb-protected H-(Ala)10-NH2) and final
product (unprotected H-(Ala)10-NH2) are shown in orange and green, respec-
tively.
We recently reported that an acceptable quality of RADA-16
can be obtained only by coupling two protected octapeptides
in solution.[42] In that study, we showed that neither the pro-
tected nor unprotected RADA-16 can be purified by chroma-
tography. We learned that the difficulty arises after the incor-
poration of 4–5 amino acids. Thus, two syntheses, a standard
for the synthesis of modified H-(Ala)10-NH2 8, which showed an
increased purity and the absence of the mentioned impurity
(Figure S8 in the Supporting Information), thus confirming an
approach to overcome the steric hindrance associated with
the N(Mmsb) residue. The main conclusion drawn from these
analyses is that the influence of the backbone protection has
an effect that is apparent even six amino acids further down
the backbone. Furthermore, in these two syntheses (8 and 9),
no peptides lacking Ala residues were detected (Figure 4b and
Figure S4b in the Supporting Information). The standard pep-
tide synthesized with common Ala (7) showed four sequences
lacking Ala residues, as shown by MS (Figure 4a); therefore,
these results are consistent with those obtained from the qual-
itative Kaiser test. In addition, the increase in solubility of the
modified sequences in common peptide solvents (H2O or H2O/
MeCN, 1:1) was extraordinary (Figure 4b).
In addition to HPLC and MS analysis of standard and modi-
fied H-(Ala)10-NH2, and by means of circular dichroism (CD), we
explored the effect on secondary structure when one Mmsb
unit was introduced into the peptide sequence (Figure 5). The
CD spectrum of the standard H-(Ala)10-NH2 peptide 7 showed
Figure 5. CD spectrum of a 25 mm solution of synthesized H-(Ala)10-NH2 in
H2O/2,2,2-trifluoroethanol (TFE) (9:1) at pH 5.2 of standard sequence 7 (solid
curve) and modified sequences 8 and 9 (dotted and dashed curves, respec-
tively).
Chem. Eur. J. 2014, 20, 15031 – 15039
www.chemeurj.org
15034
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 6. HPLC/MALDI-TOF MS/solubility assays of crude RADA-16 peptides
synthesized: a) by using the common amino acids, and b) by using the modi-
fied Fmoc-N(Mmsb)-Ala-OH in position 12. Concentration: 1 mgmLꢀ1 in H2O.
Mass impurities assigned (red label); mass expected peptide (green label).
RADA-16 (10) and a modified version (11), were carried out on
Rink amide-ChemMatrix resin,[52] by using DIPCDI/OxymaPure
as coupling reagent. On the modified peptide, Fmoc-N(Mmsb)-
Ala-OH was introduced in position 12 (11). Peptide elongations
were again checked qualitatively by the Kaiser test. Although
no double couplings were performed, the synthesis of the
standard RADA-16 showed, by Kaiser test, incomplete cou-
plings in the last four residues. Fmoc-N(Mmsb)-Ala-OH was in-
troduced as in the modified H-(Ala)10-NH2 examples. The cou-
pling of the incoming Fmoc-Asp(tBu)-OH was studied (Table 1
in the Supporting Information), concluding that the confluence
of a MeCN/DMF (3:1) solvent mixture and three couplings at
458C using DIPCDI and OxymaPure, was once again crucial for
satisfactory introduction of the incoming residue.
The final global deprotection/cleavage required prolonged
acid treatment because of the high content of Pbf side-chain
protecting groups of Arg [4 h with TFA/triisopropylsilane (TIS)/
H2O (38:1:1)]. Under these more demanding conditions, the
acid-stability of the Mmsb group was confirmed. Although no
significant differences between the two RADA-16 peptides
were detected by HPLC (Figure 6), the MALDI-TOF MS analysis
of these compounds revealed that standard RADA-16 con-
tained several deletions (Figure 6a and Figure S2 in the Sup-
porint Information; already anticipated by the Kaiser test re-
sults during synthesis), whereas the modified RADA-16 had
none. Careful study of the HPLC and MS results concluded that
the modified RADA-16 showed less than 1% of the impurity of
Figure 7. HPLC chromatograms of modified RADA-16 peptide: top) after
peptide synthesis, middle) after purification, and bottom) final peptide after
removing Mmsb group followed by desalting.
Chem. Eur. J. 2014, 20, 15031 – 15039
www.chemeurj.org
15035
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Figure 8. Analysis and characterization of synthesized modified Ab(1-42): HPLC and MALDI-TOF MS of a) crude Mmsb-protected peptide after synthesis, b) pu-
rified Mmsb-protected peptide, c) unprotected peptide after Mmsb removal, and d) solubility of Ab(1-42) at 1 mgmLꢀ1 in H2O of the Mmsb-protected ana-
logue (orange label) and final unprotected Ab(1-42) (blue label).
truncated peptide (5-mer, uncompleted incorporation of Asp
after coupling of modified N(Mmsb) residue; Figure 6b and
Figure S5 in the Supporting Information).[53] Most importantly,
the solubility of the modified RADA-16 crude peptide in water
was significantly higher than that of the standard peptide
(Figure 6).
assembly (12, 21AA), we manually inserted Fmoc-N(Mmsb)-Ala-
OH, followed by the incoming Fmoc-Phe-OH. Both amino acids
were coupled by following the same conditions described pre-
viously for other modified peptides. Once the elongation of
the peptide was complete, the peptidyl resin was treated with
TFA/TIS/H2O (38:1:1) for 2 h, again confirming the acid-stability
of Mmsb. Although the crude peptide obtained showed poor
purity (35%; Figure 8a and Figures S6 and S7 in the Support-
ing Information) by HPLC analysis on a C4 column, our princi-
pal aim was to enhance the solubility of Ab(1-42) and facilitate
its purification.
The lower solubility of the standard RADA-16, combined
with the poor HPLC resolution of the expected peptide, for
which the impurities eluted close to the desired product, pre-
vented its purification. In contrast, the greater solubility of
Mmsb-containing RADA-16 allowed its isolation. The final
modified peptide was purified by HPLC analysis on a reverse-
phase C18 column, and the Mmsb-protected peptide was iso-
lated in excellent purity (99.9%) (m/z calcd for C75H123N29O27S:
1893.89, found: 1894.69 [M+H]+; Figure 7 and Figure S9 in
the Supporting Information). To remove the Mmsb group, the
The Mmsb-containing Ab(1-42) was isolated by HPLC using
the analogous semi-preparative C4 column. The enhanced sol-
ubility of the modified peptide (Figure 8d) allowed its purifica-
tion (Figure 8b, and Figures S10 and S11 in the Supporting In-
formation). Mmsb removal was completed after 2 h, checked
by HPLC and MS (MALDI-TOF and HRMS-ESI), with both tech-
niques confirming the synthesis of the natural Ab(1-42) in 90%
purity and with the expected mass (HRMS-ESI m/z calcd for
purified peptide was treated with
a concentration of
1 mgmLꢀ1 NH4I/TFA and the backbone protecting group was
completely removed after 45 min, with only a final desalting
step (see the Supporting Information for details) required to
reach the pure target RADA-16 peptide (Figure 7 and Fig-
ure S14 in the Supporting Information).
C
212H321N55O62S2: 4693.3098, found: 1129.3273 [M+4H]+/4,
903.6643 [M+5H]+/5 and 753.3872 [M+6H]+/6; Figure 8c and
Figures S15 and S16 in the Supporting Information).
Inspired by the possibility of purifying RADA-16, we moved
onto a second level, extending the application of the Mmsb
backbone-protecting group to the Ab(1-42) sequence. This
amyloidogenic peptide was produced on an aminomethyl
resin with a previous incorporation of 3-(4-hydroxymethylphe-
noxy)propionic acid as a linker in an automatic synthesizer
with non-optimized conditions. In the middle of the sequence
Conclusion
Mmsb is a new backbone amide-protecting group that is used
in a safety-catch strategy. This new group is stable during
global deprotection conditions. Mmsb represents a new con-
cept for the synthesis and purification of “difficult peptides”. Its
Chem. Eur. J. 2014, 20, 15031 – 15039
www.chemeurj.org
15036
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
chain-protecting groups: TFA/CH2Cl2 (19:1) for 1 h (H-(Ala)10-NH2, 7)
and TFA/TIS/H2O (38:1:1) for 4 h (RADA-16, 10). The mixture was
partially evaporated under reduced pressure and the peptides
were precipitated with cold diethyl ether. The liquid layer was re-
moved by centrifugation and the solid was washed with cold di-
ethyl ether to give white solids that were dissolved in H2O/MeCN
(1:1) and lyophilized. The crude peptides were analyzed by HPLC
and MALDI-TOF MS (see Figures S1 and S2 in the Supporting
Information).
presence in a peptide confers several advantages over a stan-
dard synthesis. Thus, the elongation of the peptide sequence
takes place smoothly and with better yields because of the dis-
ruption of the secondary structure. Furthermore, its use could
prevent side-reactions, such as aspartimide and diketopipera-
zine formation, while facilitating others, such as cyclization. We
have demonstrated that the presence of only one Mmsb
molecule favors the incorporation of at least 10 amino acids in
the case of H-(Ala)10-NH2 and RADA-16. In addition, the Mmsb-
protected peptides are much more soluble than standard pep-
tides, thus allowing better HPLC resolution and, most
importantly, their purification. The improved solubility could
be explained by the absence of secondary structure and also
by the polarity of the sulfoxide. The strategy described here
could be applied to any amino acid other than Ala. Preparation
in solution of dipeptides (equivalent to pseudoprolines or dep-
sipeptides) is advisable when using Mmsb-hindered amino
acids, so as to facilitate incorporation in the peptide sequence.
The removal of Mmsb takes place under conditions that do
not damage the peptide and allows the desired peptide to be
obtained after only one filtration. Notably, our results on the
synthesis of Ab(1-42), for which only one Mmsb group was in-
troduced within 42 AA residues, support the proposed synthet-
ic approach. The results show that this novel synthetic tech-
nique enhances the manipulation of the peptide, especially
with regard to its solubility and purification. We envisage that
this technology will have broad use not only for the synthesis
of “difficult peptides”, but also for the purification of a large
range of other peptides.
Synthesis of modified peptide sequences
The modified peptide sequences [two analogues modified
H-(Ala)10-NH2: one in position 6 (8), the second in position 8 (9);
and the modified Ac-(RADA)4-NH2 in position 12 (11)] were synthe-
sized on the same resins described in case of each standard
peptide. The resin washings, commercial amino acid couplings,
and Fmoc removal cycles were performed by following the same
protocol described for the standard sequences synthetic methodol-
ogies. The Kaiser test was used to detect which amino acid was
not completely coupled (neither in 8, 9, nor 11 were uncompleted
couplings detected). No double couplings were performed except
in the case of the incoming amino acid to H-N(Mmsb)-Ala-OH. The
synthesis of modified Ab(1–42) (12) was carried out by an automat-
ic microwave peptide synthesizer (Discover, CEM corporation). The
synthesis was performed with an Aminomethyl ChemMatrix resin
(160mg, 0.1mmol, 0.62mmolgꢀ1) with a previous incorporation of
3-(4-hydroxymethylphenoxy)-propionic acid as a linker, coupled
with the same system described in standard sequences. Commer-
cial and common Fmoc-l-AA(PG)-OH amino acids were used. The
coupling system consisted of N-[(1H-benzotriazol-1-yl)-(dimethyla-
mino)methylene] N-methylmethanaminium hexafluorophosphate
(HBTU) with N,N-diisopropylethylamine (DIEA) and DMF as a solvent
and the Fmoc removal was performed with piperidine/DMF (1:4).
Experimental Section
Synthesis of standard peptide sequences
Incorporation of the synthesized building block Fmoc-
N(Mmsb)-Ala-OH
H-(Ala)10-NH2 (7) and RADA-16 (10) peptides were synthesized on
Rink-Amide AM Polystyrene resin (250 mg, 0.12 mmol,
0.48 mmolgꢀ1) and Rink-Amide AM ChemMatrix resin (100 mg,
0.053 mmol, 0.53 mmolgꢀ1), respectively. Polystyrene resin was
conditioned by washing with DMF (3ꢂ1 min) and CH2Cl2 (3ꢂ
1 min), and ChemMatrix resin was conditioned by initial washing
with TFA/CH2Cl2 (1:99) (5ꢂ1 min), CH2Cl2 (5ꢂ1 min), DIEA/CH2Cl2
(1:19) (5ꢂ1 min) and CH2Cl2 (5ꢂ1 min). Commercial amino acids
were coupled in both synthesis as follow: Fmoc-l-AA(PG)-OH
(3 equiv), DIPCDI (3 equiv), and OxymaPure (3 equiv) in DMF, with
a 5 min preactivation and with a total coupling time of 1 h. After
every coupling, the resin was washed with DMF (3ꢂ1 min) and
CH2Cl2 (3ꢂ1 min). Then, a Kaiser test was carried out to check
which amino acid was not coupled completely [4 AA: Ala1-Ala4 in
the case of H-(Ala)10-NH2 (7); and 4 AA: Ala9, Ala6, Ala4, Arg1 in the
case of RADA-16 (10)]. No double couplings were performed inde-
pendently of the test results. Next, the Fmoc removal was per-
formed with piperidine/DMF (1:4) (25 mLgꢀ1 resin, 1ꢂ1 min, 2ꢂ
5 min), followed by resin washings with DMF (3ꢂ1 min), CH2Cl2
(3ꢂ1 min), and DMF (3ꢂ1 min). The cycle of AA coupling/Fmoc re-
moval was performed until complete elongation of peptides, in-
cluding the last Fmoc removal. In the case of RADA-16 (10), a final
step of acetylation was accomplished by Ac2O (10 equiv) with DIEA
(10 equiv) in DMF for 30 min. Final resin washings were carried out
with DMF (3ꢂ1 min) and CH2Cl2 (3ꢂ1 min). The peptides were sub-
jected to cleavage treatment depending on the amino acid side
After Fmoc removal of the previous amino acid, the Fmoc-
N(Mmsb)-Ala-OH (1.5 equiv) was coupled by using DIPCDI
(1.5 equiv) and OxymaPure (1.5 equiv) in DMF for 1h. Then, wash-
ings with DMF (3ꢂ1min) and CH2Cl2 (3ꢂ1min) were performed
and Kaiser test confirmed the complete incorporation of the AA.
Incorporation of the incoming Fmoc-AA-OH to H-N(Mmsb)-
Ala-OH
Modified RADA-16 (11) was selected to find the best coupling con-
ditions to introduce the incoming amino acid (see Table S1 in the
Supporting Information) and the extension of this reaction was
evaluated by HPLC analysis of the crude peptide after an aliquot of
peptidyl-resin was treated with a cleavage cocktail. The solvent
was then evaporated and the peptide was precipitated with cold
diethyl ether. Amino acid coupling consecutive to H-N(Mmsb)-Ala-
OH was accomplished by following the same conditions used in all
modified peptide sequences, at 458C for 2h in DMF/MeCN (1:2),
with the coupling reagent system DIPCDI (10 equiv) and Oxyma-
Pure (10 equiv) for Fmoc-Phe-OH (10 equiv). A total of three
consecutive couplings (using the same coupling conditions) were
performed.
Chem. Eur. J. 2014, 20, 15031 – 15039
www.chemeurj.org
15037
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[10] R. C. de L. Milton, S. C. F. Milton, P. A. Adams, J. Am. Chem. Soc. 1990,
112, 6039–6046.
[11] J. Bedford, C. Hyde, T. Johnson, W. Jun, D. Owen, M. Quibell, R. C. Shep-
pard, Int. J. Pept. Protein Res. 1992, 40, 300–307.
Cleavage/global deprotection of side-chain protecting
groups of modified peptides
After all the peptide elongation, cleavage, and post-cleavage treat-
ments were performed by following specified conditions depend-
ing on the amino acid side chain-protecting groups nature, the
peptides were subjected to cleavage treatment TFA/CH2Cl2 (19:1)
for 1h (8 and 9), TFA/TIS/H2O (95:2.5:2.5) for 4h (11) and TFA/TIS/
H2O (95:2.5:2.5) for 2h (12). The mixture was partially evaporated
under reduced pressure and the peptides were precipitated with
cold diethyl ether. The liquid layer was removed by centrifugation
and the solid was washed with cold diethyl ether to give white
solids that were dissolved in H2O/MeCN (1:1) and lyophilized to
afford the modified sequences. The crude peptides were analyzed
by HPLC and MALDI-TOF MS. Peptides were characterized by HPLC
and MALDI-TOF MS (see Figures S3–S7 in the Supporting Infor-
mation).
[12] M. Mutter, V. N. R. Pillai, H. Anzinger, E. Bayer, C. Toniolo, in Proc. 16th
Eur. Pept. Symp. (Ed.: K. Brunfeldt), Scriptor, Copenhagen, 1981,
pp. 660–665.
[13] S. M. Meister, S. B. H. Kent, in Proc. to 8th Am. Pept. Symp. (Eds.: V. J.
Hruby, D. H. Rich), Pierce Chemical Company Co., Rockford, 1983,
pp. 103–106.
[14] T. Haack, M. Mutter, Tetrahedron Lett. 1992, 33, 1589–1592.
[15] G.-A. Cremer, H. Tariq, A. F. Delmas, J. Pept. Sci. 2006, 12, 437–442.
[16] Y. Sohma, M. Sasaki, Z. Ziora, N. Takahashi, T. Kimura, Y. Hayashi, Y. Kiso,
in Proc. to 18th Am. Pept. Symp. (Eds.: M. Chorev, T. K. Sawyer), American
Peptide Society, Boston, 2003, pp. 67–68.
[17] Y. Sohma, M. Sasaki, Y. Hayashi, T. Kimura, Y. Kiso, Chem. Commun. 2004,
124–125.
[18] L. A. Carpino, E. Krause, C. D. Sferdean, M. Schꢄmann, H. Fabian, M. Bi-
enert, M. Beyermann, Tetrahedron Lett. 2004, 45, 7519–7523.
[19] I. Coin, P. Schmieder, M. Bienert, M. Beyermann, J. Pept. Sci. 2008, 14,
299–306.
Removal of the Mmsb amide protecting group
[20] T. Johnson, M. Quibell, D. Owen, R. C. Sheppard, J. Chem. Soc. Chem.
Commun. 1993, 369–372.
[21] M. Liley, T. Johnson, Tetrahedron Lett. 2000, 41, 3983–3985.
[22] E. Nicolꢅs, M. Pujades, J. Bacardit, E. Giralt, F. Albericio, Tetrahedron Lett.
1997, 38, 2317–2320.
[23] J. Howe, M. Quibell, T. Johnson, Tetrahedron Lett. 2000, 41, 3997–4001.
[24] E. C. B. Johnson, S. B. H. Kent, Chem. Commun. 2006, 1557–1559.
[25] A. Isidro-Llobet, X. Just-Baringo, M. Alvarez, F. Albericio, Biopolymers
2008, 90, 444–449.
[26] J. P. Tam, M. W. Riemen, R. B. Merrifield, Pept. Res. 1988, 1, 6–18.
[27] E. Nicolꢅs, E. Pedroso, E. Giralt, Tetrahedron Lett. 1989, 30, 497–500.
[28] Y. Yang, W. V. Sweeney, K. Schneider, S. Thçrnqvist, B. T. Chait, J. P. Tam,
Tetrahedron Lett. 1994, 35, 9689–9692.
[29] W. Lunkenheimer, H. Zahn, Justus Liebigs Ann. Chem. 1970, 740, 1–17.
[30] B. F. Gisin, R. B. Merrifield, J. Am. Chem. Soc. 1972, 94, 3102–3106.
[31] Our definition of safety-catch protecting group is based on that used
by Pꢅtek and Lebl for the same type of linkers: “provides high stability
during the synthesis, but capable of chemoselective modification and
subsequent labilization for cleavage step”.[54]
Lyophilized modified crude peptides 8, 9, 11, and 12 (1–5 mg)
were dissolved in neat TFA (concentration 1 mgmLꢀ1) in a round-
bottomed flask with constant stirring at RT. Ammonium iodine
(30 equiv) was then added to the reaction mixture and the reaction
was monitored by HPLC (an aliquot was taken from the reaction
mixture, the TFA was removed by N2 (g), and the peptide was pre-
cipitated in diethyl ether and redissolved in H2O/MeCN (1:1)).
When MALDI-TOF MS analysis revealed the completion of the reac-
tion, removal of the iodine byproduct was carried out by filtration
of the reaction mixture. The majority of volatiles were removed
under reduced pressure and the precipitation of the peptide was
performed by transferring the peptide solution dropwise to cold
diethyl ether. After three times centrifugation and diethyl ether
washings, the solid was lyophilized and the peptides were ana-
lyzed by MALDI-TOF MS (see Figures S12–16 in the Supporting
Information).
[32] S. Thennarasu, C.-F. Liu, Tetrahedron Lett. 2010, 51, 3218–3220.
[33] J. M. Samanen, E. Brandeis, J. Org. Chem. 1988, 53, 561–569.
[34] M. Erlandsson, A. Undꢆn, Tetrahedron Lett. 2006, 47, 5829–5832.
[35] Y. Kiso, T. Kimura, M. Yoshida, J. Chem. Soc. Chem. Commun. 1989,
1511–1513.
[36] S. Lebreton, M. Pꢅtek in Linker Strategies in Solid-Phase Organic Synthe-
sis, (Ed.: P. Scott), John Wiley & Sons, New York, 2009, pp. 195-220.
[37] S. A. Kates, N. A. Sole, M. Beyermann, G. Barany, F. Albericio, Pept. Res.
1996, 9, 106–113.
[38] D. Englebretsen, P. Alewood, Tetrahedron Lett. 1996, 37, 8431-8434.
[39] S. Abdel Rahman, A. El-Kafrawy, A. Hattaba, M. F. Anwer, Amino Acids
2007, 33, 531–536.
[40] K. Hamada, M. Hirose, T. Yamashita, H. Ohgushi, J. Biomed. Mater. Res.
Part A 2008, 84, 128–136.
Acknowledgements
The work was partially supported by CICYT (CTQ2012–30930),
the Generalitat de Catalunya (2014SGR 137), and by the Insti-
tute for Research in Biomedicine (Spain). We really thank Dr.
Natalia Carulla for her helpful advice regarding the manipula-
tion of Ab(1–42).
Keywords: peptides
·
protecting groups
·
solid-phase
synthesis · synthesis design · synthetic methods
[41] Z. Luo, S. Zhang, Chem. Soc. Rev. 2012, 41, 4736–4754.
[42] M. Paradꢀs-Bas, J. Tulla-Puche, A. A. Zompra, F. Albericio, Eur. J. Org.
Chem. 2013, 5871–5878.
[1] M. Gꢃngora-Benꢀtez, J. Tulla-Puche, F. Albericio, ACS Comb. Sci. 2013, 15,
217–228.
[43] C. Haass, D. J. Selkoe, Nat. Rev. Mol. Cell Biol. 2007, 8, 101–112.
[44] S.-H. Yen, W.-K. Liu, F. L. Hall, S.-D. Yan, D. Stern, D. W. Dickson, Neuro-
biol. Aging 1995, 16, 381–387.
[45] A. B. Clippingdale, M. Macris, J. D. Wade, C. J. Barrow, J. Pept. Res. 1999,
53, 665–672.
[46] Y. Sohma, M. Sasaki, Y. Hayashi, T. Kimura, Y. Kiso, Tetrahedron Lett. 2004,
45, 5965–5968.
[47] Conditions for the alkylation and formylation steps were slightly adapt-
ed from those described in ref. [32].
[2] F. Albericio, L. A. Carpino, Methods Enzymol. 1997, 289, 104–126.
[3] S.-Y. Han, Y.-A. Kim, Tetrahedron 2004, 60, 2447–2467.
[4] T. I. Al-Warhi, H. M. A. Al-Hazimi, A. El-Faham, J. Saudi Chem. Soc. 2012,
16, 97–116.
[5] A. El-Faham, F. Albericio, Chem. Rev. 2011, 111, 6557–6602.
[6] L. A. Carpino, J. Am. Chem. Soc. 1957, 79, 98–101.
[7] A. Isidro-Llobet, M. Alvarez, F. Albericio, Chem. Rev. 2009, 109, 2455–
2504.
[8] S. B. H. Kent, in Proc. 9th Am. Pept. Symp. (Eds.: K. D. Deber, C. M. Hruby,
V. J. Kopple), Pierce Chemical Company Co., Rockford, 1985, pp. 407–
414.
[48] U. C. Dyer, D. A. Henderson, M. B. Mitchell, P. D. Tiffin, Org. Process Res.
Dev. 2002, 6, 311–316.
[49] R. Subirꢃs-Funosas, R. Prohens, R. Barbas, A. El-Faham, F. Albericio,
Chemystry - A Eur. J. 2009, 15, 9394–9403.
[9] E. Atherton, R. C. Sheppard, Solid Phase Peptide Synthesis-A Practical Ap-
proach, Oxford University Press, Oxford, England, 1989.
Chem. Eur. J. 2014, 20, 15031 – 15039
www.chemeurj.org
15038
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[50] G. A. Acosta, M. del Fresno, M. Paradis-Bas, M. Rigau-DeLlobet, S. Cꢇtꢆ,
M. Royo, F. Albericio, J. Pept. Sci. 2009, 15, 629–633.
[51] E. Kaiser, R. L. Colescott, C. D. Bossinger, P. I. Cook, Anal. Biochem. 1970,
34, 595–598.
[53] The very low percentage of 5-mer impurity showed quite different
chromatographic properties compared with the expected peptide,
which facilitates the purification step.
[54] M. Pꢅtek, M. Lebl, Biopolymers 1998, 47, 353–363.
[52] S. Cꢇtꢆ, New Polyether Based Monomers and Highly Cross-Linked Am-
phiphile Resins, 2005, WO 2005/012277 A1.
Received: May 23, 2014
Published online on October 3, 2014
Chem. Eur. J. 2014, 20, 15031 – 15039
www.chemeurj.org
15039
ꢁ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim