Angewandte
Communications
Chemie
complete the synthesis of celogentin C (Figure 1B).[13]
Despite these advances, the introduction and removal of
DGꢀs often implies additional and non-trivial steps, thus
limiting the use of C(sp3)-H activation to the modification of
building blocks prior to peptide synthesis. Recently, Yu and
co-workers[11a] postulated that peptide backbones are capable
of acting as DGs, forming a similar N,N-dicoordinated
(Ala residue in position 2 from the N-terminus) was not
observed.
In an attempt to improve the outcome of the reaction,
silver oxidants, acid additives,[15] and microwave irradiation
were screened (see the Supporting Information) but only
lower conversions were achieved. The role of the solvent
appeared to be crucial as the use of tBuOH as the only solvent
considerably suppressed side-product formation while only
incomplete conversion of the starting material (SM) was
observed (see the Supporting Information, Table S1, entry 9).
With this in mind, we investigated the effect of 1:1 solvent
mixtures. The use of toluene/tBuOH provided the best result
(88% conversion, 29% yield; see Table S1, entry 11 and
Figure S10).
With optimized reaction conditions in hand, we next
turned to the evaluation of the scope of the reaction (Table 1).
We first set out to confirm the compatibility of the reaction
conditions with natural amino acids bearing standard protect-
ing groups for Fmoc SPPS, as the pioneering work by the Yu
laboratory[11a] mainly focused on peptides featuring all-hydro-
carbon side chains. Gratifyingly, the reaction tolerated O-tert-
butyl-l-serine (Ser(tBu)), N$-(2,2,4,6,7-pentamethyldihydro-
benzofuran-5-sulfonyl)-l-arginine (Arg(Pbf)), Ne-tert-buty-
loxycarbonyl-l-lysine (Lys(Boc)), l-glutamic acid 5-tert-
butyl ester (Glu(OtBu)), and Ng-trityl-l-asparagine (Asn-
(Trt)), affording very good to excellent conversions (2b–2 f).
In addition, branched amino acids such as l-leucine (Leu) and
O-tert-butyl-l-threonine Thr(tBu) in position i + 2 did not
prevent the staple formation from proceeding with good
conversion (2g and 2h). The effect of l-isoleucine (Ile) and l-
valine (Val) on peptide stapling could not be investigated as
peptides 1i and 1j proved to be insoluble in the solvent
mixture. Unfortunately, the conformational bias resulting
from the use of l-proline (Pro) in position i + 2 (1k)
complex with the Pd catalyst. They implemented this strategy
3
À
for late-stage site-selective C(sp ) H arylation at the N-
terminal amino acid of short peptides (Figure 1C), thus
providing a practical method for rapid peptide derivatization.
Herein, we report the synthesis of a novel class of stapled
3
À
peptides based on Yuꢀs backbone-assisted C(sp ) H activa-
tion method. This new process produces an original staple
motif featuring an unprecedented C(sp ) C(sp ) linkage
between an alanine (Ala) and a phenylalanine (Phe) residue
3
2
À
(Figure 1D). This constitutes the first example of late-stage
3
À
C(sp ) H peptide macrocyclization.
We envisaged that the intermediate resulting from the Pd
3
À
À
catalyzed C H activation of the primary b-C(sp ) H bond of
phthaloyl (Phth) protected N-terminal Ala; would react
intramolecularly with an iodophenylalanine residue intro-
duced within the same peptide sequence to provide a structur-
ally unique staple. To test our hypothesis, we prepared the
linear tetrapeptide 1a using standard 9-fluorenylmethyloxy-
carbonyl solid-phase peptide synthesis (Fmoc SPPS) proce-
dures (see the Supporting Information). We then focused on
À
the key in-solution C H activation step, stirring 1a
(0.05 mmol), Pd(OAc)2 (10 mol%), and AgOAc (2 equiv) in
1,2-dichloroethane (DCE; 0.1m) at 1008C for 24 h. Pleasingly,
HPLC analysis of the crude reaction mixture showed 63%
conversion into the desired Si,i+3S(5)[14] product 2a, which was
formed as a single diastereoisomer (Figure 2). The main
impurities were identified as the deiodination and aryl–aryl
homocoupling side products. After purification, 2a was fully
characterized by NMR analysis and high-resolution mass
spectrometry. These analyses confirmed that no loss of
diastereoisomeric purity had occurred and that no cyclic
À
prevented C C bond formation. The stapling of the challeng-
ing peptides 1l/1m and 1n/1o, which bear Pd-deactivating
2
À
sulfur atoms and reactive C(sp ) H bonds, respectively, was
also investigated. Whereas S-trityl-l-cysteine (Cys(Trt)) con-
taining 1l underwent thiol elimination, the l-methionine
(Met) containing product 2m was obtained with 26%
conversion. Surprisingly, the macrocyclization of l-Trp-
À
homodimer was formed during the C H macrocyclization
step. As previously reported by Yu and co-workers,[11a] the
À
C H activation was site-selective for the N-protected termi-
À
nal Ala residue, and the product of C H activation at Ala2
(Boc)-containing 1n preferentially occurred through the
3
À
C(sp ) H bond to afford 2n with 53% conversion. Although
3
À
C(sp ) H stapling of 1-trityl-l-histidine (His(Trt)) containing
1o was observed, a complex mixture was obtained owing to
partial loss of the Trt protecting group.
Next, we focused on modifying the C-terminal amino acid
and prepared peptides 1p and 1q with 3-iodo-d-Phe and 4-
iodo-l-Phe, respectively. While changing the stereochemistry
À
did not affect the outcome of the C H arylation (2p, 88%),
the attempted cyclization of 1q, featuring the iodo substituent
in para position, led to a complex reaction mixture. We also
examined the macrocyclization of peptides 1r and 1s
containing N-terminal 2-aminoisobutyric acid (Aib) and
3
À
Phe, respectively. Interestingly, b-C(sp ) H activation of the
prochiral Aib residue in 1r occurred with desymmetrization
to afford product 2r with 57% conversion and excellent
3
À
Figure 2. Initial C(sp ) H stapling experiment. A) HPLC chromato-
gram of the pure linear peptide. B) HPLC chromatogram of the crude
cyclization reaction mixture.
diastereoselectivity (93:7 d.r.). However, the activation of the
3
À
secondary b-C(sp ) H bond of the Phe residue failed.
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 6
These are not the final page numbers!