Towards a photochemical synthesis of peptides

Article abstract of DOI:10.2533/chimia.2013.896

The laboratory preparation of peptides, once a challenge, is now a standard operation that can be automatised. However, the need for particular protecting groups requiring relatively harsh conditions for their removal (strong acids or bases) and reactive coupling agents can be problematic in specific cases (e.g. in a cell or in a microfluidic device). In this account, we describe our efforts towards a fully photochemical approach to peptide synthesis, where both the coupling and the deprotection steps use light instead of reagents. In order to selectively promote one of the two steps at a time, monochromatic light is used. Schweizerische Chemische Gesellschaft.

Full text of DOI:10.2533/chimia.2013.896

896 CHIMIA 2013, 67, Nr. 12
doi:10.2533/chimia.2013.896
PePtide Science in Switzerland
Chimia 67 (2013) 896–898 © Schweizerische Chemische Gesellschaft
Towards a Photochemical Synthesis of
Peptides
Christian G. Bochet*
Abstract: The laboratory preparation of peptides, once a challenge, is now a standard operation that can be
automatised. However, the need for particular protecting groups requiring relatively harsh conditions for their
removal (strong acids or bases) and reactive coupling agents can be problematic in specific cases (e.g. in a cell
or in a microfluidic device). In this account, we describe our efforts towards a fully photochemical approach to
peptide synthesis, where both the coupling and the deprotection steps use light instead of reagents. In order to
selectively promote one of the two steps at a time, monochromatic light is used.
Keywords: Amidation · Amides · Peptide synthesis · Photochemistry · Protecting groups
Peptides and proteins occupy a central of the Nobel Prize in chemistry to Bruce icant cross-reactivity),[8–11] we wondered
part in the chemistry of life, and it is not Merrifield in 1984. Less visible, but equal- whether we could perform both fundamen-
a surprise that their synthetic preparation ly important for the success of these meth- tal steps in peptide synthesis at two differ-
became a challenge for chemists very ear- ods, was the development of protecting ent wavelengths. If that would be the case,
ly on.[1] The intuitively rightful discon-
groups capable of surviving the synthesis then a reagentless peptide synthesis might
nective analysis of these macromolecules of the peptide, i.e. being orthogonal with become possible, since the only activator
most frequently identifies the amide bond respect to the other functional groups (this for the reaction would be just light. Thus,
as the key element, and literally hundreds term was coined by Merrifield himself).[5] we embarked on a program aiming at find-
of methods have been devised to exploit Emerging from this evolution, the acid- ing a carboxyl group activator that would
the intrinsic nucleophilicity of an amine labile Boc group[6] and the base-labile operate at a different wavelength than the
and the electrophilicity of an activated car- Fmoc group[7] have become now gold removal of the protecting group.
boxyl group to create this bond. The latter
has been augmented by various synthetic The above enthusiastic introduction
tricks, e.g. by converting the canonic OH might give the reader the impression that Photochemical Activation of a
group into a better leaving group. Thus, peptide synthesis is a solved problem, Carboxyl Group
acid chlorides or fluorides, various carbo- leaving space only to incremental improve-
standards in the field.
diimide-based activated esters, triazole de- ment of a very robust technology. This is
We initially worked on the design of
rivatives, hydroxysuccinimide derivatives, obviously not the case, and this thematic an ideal group that would be totally inert
uronium derivative, and others, have been issue clearly shows that creative research is towards nucleophilic attack, but would
introduced, solving completely or par- still able to open new horizons and to give be converted in situ into a reactive acyl
tially the problems associated with cost, to the chemist, biologist or material scien- transfer reagent upon photolysis.[12] Early
reaction rates, racemization, azlactone tist tools to address problems that appeared candidates were amidopyridine derivatives
formation, ease of purification etc.[2] The
unsolvable just a few years ago.
Based on our research on wave- structurally related photo-Fries rearrange-
such as 1 (Scheme 1).[13] Paralleling the
most significant advance was introduced
in the 1960s, with the attachment of the length-selective photochemical reactions, ment (which is well known on phenol es-
growing peptidic fragment to an insoluble more particularly in our search for chro- ters), we banked on a photoinduced migra-
polymeric support, which paved the way to matically orthogonal photosensitive pro- tion of the acyl group to the pyridine-type
automated procedures, nowadays routinely tecting groups (i.e. groups that could be nitrogen atom, leading to an intermediate
used without significant human interven- selectively removed photochemically us- reminiscent of the DMAP-accelerated
tion.[3,4] This quantum leap in the synthesis
ing a specific wavelength, without signif- acylation of alcohols.
of peptides was recognized by the award
Scheme 1. Early
attempts to devel-
op photoacylating
agents.
C H 5NH2
Me1C2 N2, 25°C
O
decomp.
hν
N
NH2
N
N
H
O
1
C H 5NH2
1
2 2
O
(CH2Cl) ,hν
N
O
2
18-47%
C H 5HN
R
1
2 2
*Correspondence: Prof. C. G. Bochet
Department of Chemistry
University of Fribourg
N
N
R
H
2a:R=Me
3a:R=Me
Chemin du Musée 9
CH-1700 Fribourg
E-mail: christian.bochet@unifr.ch
2b:R=C H
3b:R=C H
11 2
3
11 2
3

PePtide Science in Switzerland
CHIMIA 2013, 67, Nr. 12 897
However, photolysis of 1 gave a com- was published in 2001.[18,19] We used the
plex mixture of products, including, among more reactive dinitroindoline derivatives
many others, traces of the product where 4b, and studied their reaction mechanism
Photolabile Protecting Groups that
are Chromatically Orthogonal to
the Photoamidation Reaction
the acyl group indeed followed a pho- together with John Toscano in Baltimore
to-Fries rearrangement, but in the wrong by time-resolved FTIR, confirming an ear-
As the typical operating wavelength for
the indoline-based acylation lies between
350 and 420 nm, a chromatically orthogo-
nal protecting group should react either at
direction! We corrected this problem, not lier hypothesis of a photoinduced N-O acyl
too elegantly, by just flanking both ortho transfer, generating a very reactive acyl-ni-
positions with nitrogen atoms, bringing the tro intermediate, immediately trapped by
amidopyrimidines 2a,b (literally) into the a nucleophile, in general an amine.[20] We
spotlight. Despite the disappointment from later expanded the scope of the reaction to
a synthetic perspective, it was still a bitter- the formation of esters.[21]
sweet intellectual satisfaction to get a 47%
yield of acyl transfer in the irradiation of
2b in the presence of an amine. However, Preparing Amino Acid-containing
the mechanistic considerations were just Building Blocks
a longer or shorter wavelength. The former
range would pose two problems: a) no such
groups were available at that time, and b)
since the deprotection systematically oc-
curs after the coupling step, the photola-
bile group should be strictly unreactive at
the coupling wavelength, a feature that can
be delicate to attain. On the other hand, a
shorter wavelength ensures that nothing
happens during the coupling step, and,
strictly speaking, chromatic orthogonality
is not necessary: a simple modulated la-
bility is enough.[23] We chose the α,α-di-
methyl-3,5-dimethoxybenzyloxycarbonyl
(Ddz) group (Fig. 1), which has a high
photoreactivity at 300 nm and shorter.[24]
the reason for attempting the reaction; its
partial success by no means validated the
initial hypothesis.
Despite the success in triggering the
formation of amide bonds by light, we
Less creative, but massively more ef- still faced the challenge of preparing more
ficient, a careful literature search showed complex acyl donors than simple alkyl
that Patchornik et al. had developed a chains. Attempts at direct acylation of the
group that seemed to fulfill our require- very poorly nucleophilic dinitroindolines
ments: the acylated 5-bromo-7-nitroin- 5a or 5b systematically failed in our hands.
doline (abbreviated Bni, Scheme 2).[14,15] We thus went for a slightly less direct, but
Used on numerous occasions as a protect- still high yielding route: the introduction
ing group[16] or in caging compounds,[17] of the amino acid derivative in an acyclic
these nitroindoline derivatives were much precursor 6 where the nitrogen atom is not
more rarely used as acyl transfer groups, deactivated by the dinitroaromatic ring,
despite the original impressive example and then cyclize it into an indoline by an
of racemization-free synthesis of a pep- intramolecular Pd-catalyzed arene amida-
tide.[15] Unknowingly parallel work by tion (Scheme 3).[22] With this strategy, we
Nicolaou and Winssinger and us exploiting prepared a dozen different photoactivable
this reaction for the preparation of amides amino acid derivatives.
Formation of Peptides
We then tested the overall feasibility
of the scheme, by attempting to prepare a
simple dipeptide: H-Phe-Ala-OH.[25] The
coupling step proceeded smoothly at 385
nm, as well as the deprotection at 300 nm.
In order to check for a potential epimer-
ization, an unwanted side reaction some-
times observed in peptide coupling, we
prepared the diastereomeric pair of dipep-
tides 12a and 12b, starting with l-Ala-Dni
10 and its enantiomer d-Ala-Dni ent-10,
and no cross-contamination was observed
(Scheme 4).
X
X
+
X
hν
NuH
O
N
N
H
N
R
Nu
R
C H
1 23
NO2
1
NO2
N
O
O
O
O
5a:X=Br (Bni)
4a: X=Br
4b:X=N 2
5b:X=N 2
O (Dni)
With this result in hand, we embarked
O
on the synthesis of a full pentapeptide
(Scheme 5), the OGP(10,14),[26] which is
the active portion of an osteogenic growth
protein (we chose it based on the library
of photoactivable amino acids that we had
in hand, and not for its biological proper-
ties). Thus, exposure to 385 nm light of a
mixture of glycine (the starting amino acid
of the peptide) and photoactivable glycine
(the next building block) smoothly led to
the protected dipeptide, which was then
deprotected by exposure to 300 nm light.
Scheme 2. Nitroindoline derivatives as efficient photoacylating agents.
Pd(dba)
2
Me-Phos
K2CO3
O N
Z-Leu
2
H
N
IB
C
D, N
MM
O N
O N
NH2
2
2
NHZ
µ
w, 100°C
94%
NEt3
N
O
HCl
N 2
O
86%
Br
Br
O
N
H
NO2
N 2
O
8
Z
7
6
Scheme 3. Preparation of photoactivable amino acid derivatives. IBCD = iBuOCOCl; NMM =
N-methyl-morpholine.
O N
2
hν (385 nm)
N
t
t
11
Ddz-L-Ala-L-Phe-O Bu
Ddz-L-Ala-Dni + L-Phe-O Bu.HCl
N 2
O
MeCN/NEt
3
O
10
12a
17h, 84%
NH
O
O
hν (385 nm)
O
Me
t
t
Ddz-D-Ala-L-Phe-O Bu
Ddz-D-Ala-Dni + L-Phe-O Bu.HCl
11
MeCN/NEt
3
ent-10
12b
17h, 80%
9
O
Me
Fig. 1. Ddz-amino acid derivative.
Scheme 4. Epimerization-free formation of dipeptides.

898 CHIMIA 2013, 67, Nr. 12
PePtide Science in Switzerland
Scheme 5. General
strategy for the syn-
thesis of a pentapep-
tide.
Iterations of the same sequence of steps
allowed us to isolate the protected penta-
peptide, without the use of reagents other
than each building block (Scheme 6).
Conclusion
This work constitutes a proof of prin-
ciple for a full photochemical synthesis of
peptides. Of course, several limitations re-
DniꢀAmino acidꢀDdz
main, among them the tedious and costly
preparation of the building blocks, making
the overall strategy at this stage far less at-
O
O
H
N
H
N
hν (3
8
5 nm)
hν (3
0
0 nm)
t
tractive than the classical Merrifield-type
O
Bu
Ddz
Dni
Ddz
t
t
+
H N
N
H
N
O Bu
2
H N
O Bu
2
·HCl
Et N,Me
CN
H
Me
CN
synthesis. To replace that synthesis was
3 9
3
%
O
O
O
O
74%
actually never our intention. On the other
hand, wherever classical methods cannot
be applied (microfluidic devices, living
cells,...), variations of this approach could
be considered and specific applications
optimized. Photochemical reactions have
been carried out in living organisms, such
as zebrafish.[27]
hν (3
85 nm)
O
O
O
O
hν (385 nm)
H
N
H
N
H
N
Ddz-Phe-Dni
Ddz-Gly-Dni
hν (3
00 nm)
N
H N
2
t
t
N
O Bu
Ddz
N
O Bu
MeCN
Me
C
H
MeCN
H
O
O
87%
80%
90%
Ph
Ph
O
O
O
O
hν (3
8
5 nm)
H
N
H
H
N
H
t
hν (3
Me
0
0 nm)
N
Ddz-Tyr( Bu)-Dni
Ddz
N
t
t
H N
2
N
H
O Bu
N
H
N
O Bu
C
N
MeCN
H
O
O
O
O
45
%
75%
Ph
Ph
O
O
O
Acknowledgement
H
H
N
H
N
N
O
O
O N
2
t
Ddz
N
N
O Bu
The author is deeply indebted to all the
H
H
O
Me
Dni =
Ddz =
O
O
coworkers who contributed directly or indirect-
ly to this project, and whose names are men-
tioned in the original publications, in particu-
lar Céline Helgen, Aurélien Blanc, Jean-Luc
Débieux and Anne Cosandey.
Ph
N
NO2
t
Bu
O
O
Me
Scheme 6. Full photochemical synthesis of OGP(10,14).
Received: September 28, 2013
[22] J.-L. Débieux, C. G. Bochet, J. Org. Chem.
2009, 74, 4519.
[12] J.-L. Débieux, C. G. Bochet, J. Phys. Org.
Chem. 2010, 23, 272.
[1] E. Fischer, Berichte der deutschen chemischen
Gesellschaft 1903, 36, 2094.
[2] A. El-Faham, F. Albericio, Chem. Rev. 2011,
111, 6557.
[3] R. B. Merrifield, Science 1986, 232, 341.
[4] R. B. Merrifield, Science 1965, 150, 178.
[5] R. B. Merrifield, G. Barany, W. L. Cosand, M.
[13] C. Helgen, C. G. Bochet, Heterocycles 2006, [23] M. Schelhaas, H. Waldmann, Angew. Chem. Int.
Ed. 1996, 35, 2056.
67, 797.
[14] B. Amit, D. A. Ben-Efraim, A. Patchornik, J. [24] C. Birr, W. Lochinger, G. Stahnke, P. Lang,
Justus Liebigs Ann. Chem. 1972, 763, 162.
Am. Chem. Soc. 1976, 98, 843.
[15] S. Pass, B. Amit, A. Patchornik, J. Am. Chem. [25] J.-L. Débieux, C. G. Bochet, Chem. Sci. 2012,
3, 405.
Soc. 1981, 103, 7674.
Engelhard, S. Mojsov, Pept., Proc. Am. Pept. [16] G. Goissis, B. W. Erickson, R. B. Merrifield, [26] A. Spreafico, B. Frediani, C. Capperucci, A.
Symp. 5th 1977, 488.
Pept., Proc. Am. Pept. Symp., 5th 1977, 559.
Leonini, D. Gambera, P. Ferrata, S. Rosini, A.
[6] G. W. Anderson, A. C. McGregor, J. Am. Chem. [17] J. Morrison, P. Wan, J. E. T.
Corrie, G.
Di Stefano, M. Galeazzi, R. Marcolongo, J.
Cell. Biochem. 2006, 98, 1007.
Papageorgiou, Photochem. Photobiol. Sci.
Soc. 1957, 79, 6180.
[7] L. A. Carpino, G. Y. Han, J. Org. Chem. 1972,
[27] D. K. Sinha, P. Neveu, N. Gagey, I. Aujard, C.
2002, 1, 960.
37, 3404.
[18] K. C. Nicolaou, B. S. Safina, N. Winssinger,
Synlett 2001, 900.
Benbrahim-Bouzidi, T. Le Saux, C. Rampon,
[8] C. G. Bochet, Synlett 2004, 2268.
[9] A. Blanc, C. G. Bochet, J. Org. Chem. 2002, 67, [19] C. Helgen, C. G. Bochet, Synlett 2001, 1968.
C. Gauron, B. Goetz, S. Dubruille, M. Baaden,
M. Volovitch, D. Bensimon, S. Vriz, L. Jullien,
ChemBioChem 2010, 11, 653.
5567.
[20] A. D. Cohen, C. Helgen, C. G. Bochet, J. P.
[10] C. G. Bochet, Tetrahedron Lett. 2000, 41, 6341.
Toscano, Org. Lett. 2005, 7, 2845.
[11] C. G. Bochet, Angew. Chem. Int. Ed. 2001, 40, [21] J.-L. Débieux, A. Cosandey, C. Helgen, C. G.
2071.
Bochet, Eur. J. Org. Chem. 2007, 2073.