Diverse organo-peptide macrocycles via a fast and catalyst-free oxime/intein-mediated dual ligation

Article abstract of DOI:10.1039/c1cc13533c

Macrocyclic Organo-Peptide Hybrids (MOrPHs) can be prepared from genetically encoded polypeptides via a chemoselective and catalyst-free reaction between a trifunctional oxyamino/amino-thiol synthetic precursor and an intein-fusion protein incorporating a bioorthogonal keto group.

Full text of DOI:10.1039/c1cc13533c

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2012, 48, 1461–1463
www.rsc.org/chemcomm
COMMUNICATION
Diverse organo-peptide macrocycles via a fast and catalyst-free
oxime/intein-mediated dual ligationwz
Maragani Satyanarayana, Francesca Vitali, John R. Frost and Rudi Fasan*
Received 15th June 2011, Accepted 11th August 2011
DOI: 10.1039/c1cc13533c
Macrocyclic Organo-Peptide Hybrids (MOrPHs) can be prepared
from genetically encoded polypeptides via a chemoselective and
catalyst-free reaction between a trifunctional oxyamino/amino-thiol
synthetic precursor and an intein-fusion protein incorporating a
bioorthogonal keto group.
between the unnatural amino acid para-acetylphenylalanine
(pAcF)7 and an intein (species ‘a’ in Fig. 1A). This protein would
display two functional groups with orthogonal reactivity, namely
the keto group of pAcF at the N-terminus of the target sequence
and the reactive thioester bond transiently formed at its
C-terminus via intein-catalyzed N,S-acyl transfer (species ‘b’).
Macrocyclization could then be achieved via a synthetic precursor
equipped with (i) an oxyamino group to form a stable oxime
linkage with a pAcF side chain, and (ii) an amino-thiol moiety to
coordinate an intein-mediated ligation and concomitant excision
of the intein from the biosynthetic precursor.
Macrocyclic peptides and peptide-based structures have
attracted significant interest as a source of chemical probes
and therapeutic agents.1 While peptides and peptidomimetics
in rigidified configurations can be prepared synthetically,2
genetic encoding offers the advantage to couple the creation of
vast chemical libraries (107–1010) with ultrahigh-throughput
screening methods.3–5 Notable approaches involve the introduction
of disulfide bridges within randomized peptide sequences3 or
formation of cyclic peptides via split intein-mediated cyclization,
but the range of building blocks available to assemble these
structures remains inherently limited compared to synthetic
methods.4 Alternatively, ribosomal peptides have been
constrained through the use of cysteine- or amine-reactive
cross-linking agents but these methods rely on non-directional
and non-bioorthogonal reactions which limits the choice of the
cross-linking scaffolds and it may lead to multiple undesired
products.5 To overcome these major limitations, we have
undertaken efforts toward implementing general methods for
chemoselectively embedding variable synthetic scaffolds within
ribosomal peptides to generate macrocycles with a hybrid
peptidic/non-peptidic backbone, referred to as Macrocyclic
Organo-Peptide Hybrids or MOrPHs.6 Here, we report an
efficient strategy for MOrPH synthesis which exploits a highly
chemoselective, bioorthogonal, and catalyst-free tandem
reaction between a trifunctional oxyamino/amino-thiol synthetic
precursor (SP) and genetically encoded biosynthetic precursors
(BPs) incorporating a keto group (Fig. 1A).
To test this approach, we prepared a first set of six biosynthetic
precursors with target sequences spanning 4, 5, 6, 8, 10, or 12
amino acids (CBD4(pAcF) to CBD12(pAcF), Table S1 (ESIz)).
Based on our recent investigations,6 we envisioned that a
suitable biosynthetic precursor for MOrPH construction could
be generated by framing a target peptide sequence (‘TS’)
Department of Chemistry, University of Rochester, Rochester NY,
USA. E-mail: fasan@chem.rochester.edu; Fax: +1 585-276-0205;
Tel: +1 585-275-3504
w This article is part of the ChemComm. ‘Emerging Investigators
2012’ themed issue.
z Electronic supplementary information (ESI) available: Experimental
details, synthetic procedures, characterization data, and supplemental
figures and tables. See DOI: 10.1039/c1cc13533c
Fig. 1 (A) Synthesis of Macrocyclic Organo-Peptide Hybrids via
oxime/intein-mediated tandem ligation. CBD: Chitin Binding Domain.
TAG: amber stop codon. TS: Target Sequence. GyrA: intein GyrA
from Mycobacterium xenopi. (B) Oxyamine/amino-thiol synthetic
precursors. Their synthesis is described in Schemes S2–S5 of ESI.z
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 1461–1463 1461

View Article Online
A Chitin Binding Domain (CBD, 71 amino acids) was introduced
as a N-terminal tail to mimic the molecular arrangement in a
typical display system, where a variable amino acid sequence is
tethered to a viral or cellular protein. pAcF was synthesized in
three steps from 4-methyl-acetophenone via an optimized
version of a described procedure7 (overall yield: 86%, Scheme
S1 (ESIz)). The six precursor proteins were expressed in E. coli
cells co-tranformed with a plasmid encoding for the precursor
protein (pBP) and one containing a tRNACUA/pAcF-tRNA
synthetase pair (pEVOL-pAcF)7 for suppressing the amber stop
codon (TAG) located upstream of the target sequence with pAcF.
Hydrazino, hydrazido, and oxyamino groups were all viable
choices to mediate coupling of the synthetic precursor to the
side-chain ketone in the precursor proteins. In preliminary
and the acyclic SP-bound product (Fig. S2–S3, ESIz). In stark
contrast, the reactions with 3 exhibited a considerably larger
extent of protein splicing (50–80%) after 5 hours (Fig. 2A and
Fig. S1 (ESIz)) and quantitative splicing (90–100%) after
overnight. Remarkably, MALDI-TOF analysis showed the
formation of the desired CBD-tethered hybrid macrocycle as
the only product from these reactions (Fig. 2B). Also notable
was that 3-induced macrocyclization occurred with almost
equally high efficiency across all the different target sequence
lengths, including the short 4mer (TGST) and the considerably
longer 12mer sequence (TGSWGKLAEYGT). Furthermore,
whereas the biosynthetic precursors undergo slow hydrolysis
(5–8% after 5 hours, Fig. 2A), the observation of no or
minimal amounts of hydrolyzed product (‘h’) in the presence of
3 (Fig. 2B) indicated that the macrocyclization largely outcompetes
the undesired hydrolysis process, even in the context of the
extended 10- and 12-amino acid target sequences.
experiments,
a GyrA-fused protein was incubated with
methylhydrazine, phenylhydrazide or methoxyamine at 50 mM
in phosphate buffer (pH 7.5). Methoxyamine was unable to induce
intein splicing at detectable levels even after overnight incubation
at room temperature which suggested that oxyamino-containing
SPs would not react with the thioester bond through the
nucleophilic –ONH2 group, providing the desired orthogonality
between these functional groups. Oxime linkages also exhibit
considerably higher hydrolytic stability compared to hydrazones.8
Based on work in the area of Native Chemical Ligation (NCL)
and its extensions,9,10 we suspected that the pKa, nucleophilicity,
and spatial arrangement of the amino and thiol groups could
influence considerably the ability of the SP amino-thiol moiety to
undergo the desired transthioesterification and S-N acyl
transfer reactions. Accordingly, three different trifunctional
SPs were synthesized (Fig. 1B, Schemes S2–S5 (ESIz)). Given
the reactivity of intein-fusion proteins toward peptides with
N-terminal cysteines in Expressed Protein Ligation (EPL),11,12
SP1 (1) was prepared based on this amino acid, which features
a 1,2-amino-thiol connectivity and an alkyl thiol with a pKa of
B8.5. SP2 (2) was designed to carry a more acidic thiol (pKa of
o-amino-thiophenol E 6.6)13 and present the same 1,2-amino-thiol
connectivity as 1, but rigidified through installation onto an
aromatic ring. The third design, SP3 (3), integrates a benzylic
thiol (pKa E 9.5)14 in a semi-rigid configuration and
1,3-arrangement with respect to the amino group.
To establish whether the o-amino-benzyl-thiol moiety of
3 was capable of undergoing the desired S-N acyl transfer
after transthioesterification, the reaction mixtures were treated
with iodoacetamide (20 mM) for 2 hours followed by MALDI-
TOF analysis. Such treatment led to complete disappearance of
the [M + H]+ species corresponding to the MOrPHs and to
the appearance of a species with an m/z of +58 (Fig. S4 (ESIz)),
which is consistent with the addition of an acetamido moiety to
these molecules. These tests evidenced the accessibility of the
benzylic thiol in the macrocycles to attack by the alkylating agent,
supporting the occurrence of the amide-forming intramolecular
rearrangement.
Given the identical alkoxyamino group and linker connecting
this group to the amino-thiol moiety, 1, 2, and 3 are expected to
Next, we tested the feasibility of the strategy outlined in
Fig. 1A by performing reactions where each protein construct
(100 mM) was exposed to 1, 2, or 3 (15 mM) in phosphate
buffer at pH 7.5. Tris(2-carboxyethyl)phosphine (TCEP,
20 mM) was added to the reaction mixtures to maintain the
thiol groups in the reactants in reduced form. Splicing of the
protein constructs over time was quantified by SDS-PAGE
and densitometric analysis of the gel bands corresponding to
the full-length protein (31 kDa) and the splicing fragments,
GyrA (22 kDa) and the CBD-linked products (8 kDa)
(Fig. S1, ESIz). These experiments revealed that 1 and 2 were
poorly efficient in promoting splicing of the GyrA intein across
all the protein constructs (Fig. 2A). After 5 hours at room
temperature, 1- and 2-induced splicing of the protein precursors
ranged from 5 to 15%, which was comparable to that of the
negative controls with no synthetic precursor. In addition,
MALDI-TOF analysis showed no trace of the desired
macrocycles, while observed species corresponded to the
hydrolysis product CBD-(pAcF)-(target sequence)-COOH
Fig. 2 (A) Percentage of precursor protein splicing in the presence of
no SP, 1, 2, or 3 after 5 hours as determined by SDS-PAGE. Error bars
are from triplicate experiments. Protein constructs are described in
Table S1 (ESIz). (B) MALDI-TOF spectra of MOrPHs obtained from
reaction of 3 with precursor proteins CBD4(pAcF) to CBD12(pAcF).
Calculated (c) and observed (o) m/z values corresponding to the
[M + H]+ adduct of the macrocyclic product (‘m’) are indicated.
c
1462 Chem. Commun., 2012, 48, 1461–1463
This journal is The Royal Society of Chemistry 2012

View Article Online
possess the same reactivity toward oxime formation. Thus, their
differential performance in promoting MOrPH formation
(Fig. 2, Fig. S2, S3 (ESIz)) can be attributed to the intein
splicing properties of their amino-thiol moieties. The poor
performance of the cysteine-based 1 can be rationalized con-
sidering the stringency of the applied conditions (SP at 15 mM,
no thiol catalyst added, short incubation time) compared to
EPL protocols, which typically involve high concentrations of
thiol catalysts (up to 200 mM) as well as longer reaction times.11
More surprising was the inefficiency of 2 to induce intein
splicing given that thiophenol and related aromatic thiols,
including 4-aminothiophenol, are effective catalysts for NCL
reactions.10 We conclude that the ortho amino group drastically
reduces the nucleophilicity of the neighboring thiol in the
context of intein splicing, possibly due to steric effects and/or
unfavorable hydrogen bonding interactions with the protein. By
comparison, the MOrPH-forming ability of 3 stems from the
superior intein splicing properties of its 2-amino-benzylthiol, a
structure which has never been described in the context of
thioester- or intein-mediated ligations.15 Clearly, such a struc-
ture preserves the nucleophilicity of the benzylic thiol while
placing the amino group at a viable distance for acyl transfer via
a six-membered ring intermediate.
Fig. 3 Extent of 3-induced protein splicing for 18 variants from the
library of precursor proteins with randomized 5mer (A) and 8mer (B)
target sequences. See also Tables S2 and S3 in ESI.z
Science Foundation grant CHE-0840410. We thank Peter G.
Schultz for kindly providing pEVOL_pAcF vector.
Notes and references
1 E. M. Driggers, S. P. Hale, J. Lee and N. K. Terrett, Nat. Rev.
Drug Discovery, 2008, 7, 608–624.
To investigate the possibility of diversifying the macrocycle
structures by varying their genetically encoded moiety, we
constructed two biosynthetic precursor libraries with randomized
5mer and 8mer target sequences, namely CBD-(pAcF)-X4T-GyrA
and CBD-(pAcF)-X7T-GyrA, where X corresponds to a fully
randomized position (NNK codon). About 5000 recombinants
from each library were pooled together and expressed in
E. coli. SDS-PAGE revealed only small amounts of premature
splicing during expression (o15–20%). For both libraries,
3 induced more than 35% and 60% splicing of the full-length
proteins after 5 hours and 16 hours, respectively. To establish
the occurrence of macrocyclization, 18 randomly chosen
recombinants from each library were isolated and characterized.
Remarkably, all the recombinants from the 5mer BP library and
all but one of the 18 recombinants from the 8mer BP library
yielded the desired hybrid macrocycle (Tables S2 and S3,
ESIz). For only 2/18 of the 5mer BPs and 1/18 of the 8mer
BPs a small amount of acyclic product (15–25%) was observed.
Notably, the majority of the 5mer and 8mer BP variants (63% and
58%, respectively) underwent more than 50% splicing after
overnight incubation at room temperature (Fig. 3). Most
importantly, these experiments proved the functionality of the
method across largely divergent target sequences and demonstrated
its versatility in generating diversified MOrPH structures.
2 J. M. Humphrey and A. R. Chamberlin, Chem. Rev., 1997, 97,
2243–2266; P. Li, P. P. Roller and J. Xu, Curr. Org. Chem., 2002, 6,
411–440; R. Fasan, R. L. Dias, K. Moehle, O. Zerbe, D. Obrecht,
P. R. Mittl, M. G. Grutter and J. A. Robinson, ChemBioChem,
2006, 7, 515–526; G. T. Bourne, J. L. Nielson, J. F. Coughlan,
P. Darwen, M. R. Campitelli, D. A. Horton, A. Rhumann,
S. G. Love, T. T. Tran and M. L. Smythe, Methods Mol. Biol.,
2005, 298, 151–165; V. S. Fluxa and J. L. Reymond, Bioorg. Med.
Chem., 2009, 17, 1018–1025.
3 K. T. O’Neil, R. H. Hoess, S. A. Jackson, N. S. Ramachandran,
S. A. Mousa and W. F. DeGrado, Proteins: Struct., Funct., Genet.,
1992, 14, 509–515; W. L. DeLano, M. H. Ultsch, A. M. de Vos and
J. A. Wells, Science, 2000, 287, 1279–1283.
4 C. P. Scott, E. Abel-Santos, M. Wall, D. C. Wahnon and
S. J. Benkovic, Proc. Natl. Acad. Sci. U. S. A., 1999, 96, 13638–13643.
5 S. W. Millward, S. Fiacco, R. J. Austin and R. W. Roberts, ACS
Chem. Biol., 2007, 2, 625–634; T. Kawakami, A. Ohta, M. Ohuchi,
H. Ashigai, H. Murakami and H. Suga, Nat. Chem. Biol., 2009, 5,
888–890; C. Heinis, T. Rutherford, S. Freund and G. Winter, Nat.
Chem. Biol., 2009, 5, 502–507.
6 J. M. Smith, F. Vitali, S. A. Archer and R. Fasan, Angew. Chem.,
Int. Ed., 2011, 50, 5075–5080.
7 L. Wang, Z. Zhang, A. Brock and P. G. Schultz, Proc. Natl. Acad.
Sci. U. S. A., 2003, 100, 56–61.
8 J. Kalia and R. T. Raines, Angew. Chem., Int. Ed., 2008, 47,
7523–7526.
9 P. E. Dawson, T. W. Muir, I. Clark-Lewis and S. B. Kent, Science,
1994, 266, 776–779; J. Offer and P. E. Dawson, Org. Lett., 2000, 2,
23–26; D. L. J. Clive, S. Hisaindee and D. M. Coltart, J. Org.
Chem., 2003, 68, 9247–9254; G. Chen, J. D. Warren, J. H. Chen,
B. Wu, Q. Wan and S. J. Danishefsky, J. Am. Chem. Soc., 2006,
128, 7460–7462.
10 L. E. Canne, S. J. Bark and S. B. H. Kent, J. Am. Chem. Soc., 1996,
118, 5891–5896.
11 T. W. Muir, D. Sondhi and P. A. Cole, Proc. Natl. Acad. Sci.
U. S. A., 1998, 95, 6705–6710; T. C. Evans, J. Benner and
M. Q. Xu, Protein Sci., 1998, 7, 2256–2264.
In summary, we have developed an efficient method to
construct Macrocyclic Organo-Peptide Hybrids via a dual
oxime/intein-mediated ligation. The chemoselectivity,
bioorthogonality and catalyst-free nature of this strategy and
its demonstrated efficiency in the context of precursor target
sequences of varying length and randomized composition hold
promise toward exploiting it to generate diversified MOrPHs
tethered to a viral/cellular surface of a display system. Efforts
are ongoing to investigate this approach toward the isolation
of MOrPH-based ligands for selective protein recognition.
This work was supported by startup funds from the University
of Rochester. MS instrumentation was supported by National
12 J. A. Camarero and T. W. Muir, J. Am. Chem. Soc., 1999, 121,
5597–5598.
13 J. P. Danehy and C. J. Noel, J. Am. Chem. Soc., 1960, 82, 2511–2515.
14 M. M. Kreevoy, E. T. Harper, R. E. Duvall, H. S. Wilgus and
L. T. Ditsch, J. Am. Chem. Soc., 1960, 82, 4899–4902.
15 S. Chattopadhaya, F. B. Abu Bakar and S. Q. Yao, Methods
Enzymol., 2009, 462, 195–223.
c
This journal is The Royal Society of Chemistry 2012
Chem. Commun., 2012, 48, 1461–1463 1463