Peptide macrocyclization through amide-to-amide transpeptidation

Article abstract of DOI:10.1016/j.tet.2014.05.112

Cyclic cysteine peptides are peptide macrocycles endowed with enhanced metabolic stability and potentially, with membrane permeability. They have attracted attention in drug design and interest in their synthesis. The chemical approach for macrocyclization through transpeptidation bears striking similarity to the biological approach using an intein. Both use a similar design of thioester precursors and an amide-to-amide transpeptidation scheme, employing a series of acyl shifts to break and make amide bonds. Here we describe the synthesis of two cyclic cysteine peptides, hedyotide B1 and sunflower trypsin inhibitor-1, highlighting the similarities between the intein-based and chemical amide-to-amide schemes. In our intein-based and chemical schemes, we employed an intein Mxe or a thioethylbutylamido linkage at the C-terminus of their linear precursors, respectively. Our results demonstrated that the chemical approach provides a useful alternative to the intein approach with high efficiency.

Full text of DOI:10.1016/j.tet.2014.05.112

Accepted Manuscript
Peptide macrocyclization through amide-to-amide transpeptidation
Xinya Hemu , Yibo Qiu , James P. Tam
PII:
S0040-4020(14)00842-4
10.1016/j.tet.2014.05.112
TET 25661
DOI:
Reference:
To appear in: Tetrahedron
Received Date: 21 April 2014
Revised Date: 28 May 2014
Accepted Date: 30 May 2014
Please cite this article as: Hemu X, Qiu Y, Tam JP, Peptide macrocyclization through amide-to-amide
transpeptidation, Tetrahedron (2014), doi: 10.1016/j.tet.2014.05.112.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Graphical Abstract
To create your abstract, type over the instructions in the template box below.
Fonts or abstract dimensions should not be changed or altered.
Peptide macrocyclization through amide-to-
amide transpeptidation
Leave this area blank for abstract info.
Xinya Hemu, Yibo Qiu, James P. Tam
School of Biological Sciences, Nanyang Technological University, 60 Nanyang Drive, 03s-71, Singapore
637551

1
ACCEPTED MANUSCRIPT
Tetrahedron
journal homepage: www.elsevier.com
Peptide macrocyclization through amide-to-amide transpeptidation
Xinya Hemua , Yibo Qiua, and James P. Tama
a School of Biological Sciences, Nanyang Technological University, 60 Nanyang Drive, 03s-71, Singapore 63755
ARTICLE INFO
ABSTRACT
Article history:
Received
Received in revised form
Accepted
Cyclic cysteine peptides are peptide macrocycles endowed with enhanced metabolic stability
and potentially, with membrane permeability. They have attracted attention in drug design and
interest in their synthesis. The chemical approach for macrocyclization through transpeptidation
bears striking similarity to the biological approach using an intein. Both use a similar design of
thioester precursors and an amide-to-amide transpeptidation scheme, employing a series of acyl
shifts to break and make amide bonds. Here we describe the synthesis of two cyclic cysteine
peptides, hedyotide B1 and sunflower trypsin inhibitor-1, highlighting the similarities between
the intein-based and chemical amide-to-amide schemes. In our intein-based and chemical
schemes, we employed an intein Mxe or a thioethylbutylamido linkage at the C-terminus of their
linear precursors, respectively. Our results demonstrated that the chemical approach provides a
useful alternative to the intein approach with high efficiency. (136 words, 997 characters)
Available online
Keywords:
Macrocyclization
Amide-to-amide synthesis
Transpeptidation
Amido thioester surrogate
Cys-thioester ligation
2009 Elsevier Ltd. All rights reserved
.
———
Corresponding author. Tel.: +65-63162863; fax: +65-65151632; e-mail: jptam@ntu.edu.sg

2
Tetrahedron
ACCEPTED MANUSCRIPT
Fmoc chemistry.39 Under acidic conditions, N-alkyl cysteinyl
1. Introduction
amide readily undergoes an N-S acyl shift from an amide to a
thioester at a much faster rate than the cysteinyl amide bond.40
Various simplified versions of N-alkyl cysteine have also been
developed41,42 and an example is the C-terminal
thioethylbutylamido (TEBA) linker recently introduced by our
laboratory.22 An advantage of the TEBA linker compared to a
MeCys moiety is that it does not contain an α-carbonyl group.
And eliminating the carbonyl group in TEBA also eliminates the
β-elimination side reaction of the MeCys in peptide synthesis
using Fmoc chemistry.21
Ribosome-derived,
end-to-end
macrocyclic
peptides
commonly occur in nature. They include pilins1 and circular
bacteriocins2 from bacteria, θ-defensins3 from animals,
cyclotides4,5 and sunflower trypsin inhibitors6 from plants. The
cyclized backbone protects these peptides from degradation by
exopeptidases.7 In addition, cyclic cysteine-rich peptides
possessing one or more disulfide bonds further enhance their
resistance to degradation by heat, chemical or enzymatic
treatment.8,9 Certain cyclic cysteine-rich peptides such as
cyclotides and SFTI-1 are cell membrane permeable and orally
active.10,11 Consequently, naturally-occurring cyclic cysteine-
containing peptides provide useful scaffolds as templates for
peptidyl drug development.12,13 Our laboratory has a strong
interest in the discovery of cysteine-rich peptides, both linear and
cyclic forms, from medicinal plants14,15 in addition to methods for
their synthesis.16-22
In the bioprocessing of
a
ribosome-derived peptide
macrocycle, a cyclase is required for transpeptidation reactions,
catalyzing the cleavage of a mature domain from its precursor
and the formation of an end-to-end peptide bond by joining their
terminal residues. However, only three cyclases, TraF, PATG
and PCY1, have been isolated and characterized,23-25 but not been
exploited for chemical synthesis of peptide macrocyclization.
Cyclases mediating the cleavage and cyclization reactions for θ-
defensins and cyclotides remain to be discovered. Asparaginyl
endopeptidase (AEP) has been implicated in the cleavage and
cyclization of cyclotides, but its isolation and characterization
remain elusive.26 However, intein and Sortase A, two
transpeptidases which are not naturally-occurring cyclases, have
been exploited successfully for peptide macrocyclization.27-30
Scheme 1. Schematic illustration of intein-catalyzed synthesis and
chemical approaches using thioester surrogates for preparing
macrocyclic peptides.
For its application in macrocyclization, the TEBA linker is
placed at the C-terminus of a linear construct Cys-peptide-TEBA
with Cys at the N-terminus (Scheme 2). Such a construct is
similar in arrangement to the precursor construct of Cys-peptide-
intein in which intein mediates the cyclization reaction. The
TEBA linker also mediates macrocyclization of a Cys-peptide-
TEBA construct. Under acidic conditions, the TEBA moiety
isomerizes via an N-S acyl shift reaction from an amide to a
thioester, which is trapped by an N-terminal Cys-thiol via a thiol-
thioester exchange, an S-S acyl shift reaction, to form an end-to-
end thiolactone. Interestingly, such thiolactone is relatively stable
under acidic conditions as shown by our recent synthesis of
cyclic conotoxin21 and kalata B1.22 Under basic conditions, the
end-to-end thiolactone undergoes a proximity-driven S-N acyl
shift reaction, resulting in the formation of a new amide bond to
yield a macrocycle. Advantages of the TEBA linkage as a
thioester surrogate include its ease of synthesis and compatibility
with Fmoc chemistry.
Intein, embedded as an endopeptidase in a protein, is a
protein-splicing enzyme that mediates self-excision and ligation
of the two flanking segments, N- and C-extein, to produce a
newly joined N-C extein as a mature sequence.31-33 Such splicing
process catalyzed by intein involves four-step acyl shift
reactions.34,35 The first step that initiates the chain of acyl shifts is
the N-S/O acyl shift, which breaks the amide bond between N-
extein and N-terminal Cys/Ser of intein and forms an ester or a
thioester in the catalytic pocket. This step is facilitated by a cis-
conformation of the scissile amide bond. The second acyl shift
reaction is an S/O-S/O transesterification reaction by the
nucleophilic attack from N-terminal Cys/Ser of C-extein joining
N- and C-exteins as a branched peptide via an ester or a thioester
bond. The third step involves intein self-excision through an N-N
acyl shift reaction that occurs at C-terminal Asn of intein via
succinimide formation. The final step is not enzyme catalyzed
and occurs spontaneously via an S/O-N acyl shift reaction that
joins the N- and C-exteins as a new peptide bond, a mechanism
identical to Native Chemical Ligation.36,37 Muir’s group has
successfully exploited the intein biosynthesis for protein
engineering in expressed protein ligation (EPL).38 Since then
intein-splicing techniques have been applied for ligation,
cyclization and protein labeling. For preparing cyclic peptides
using a recombinant system, a useful construct is a linear
precursor design of Cys-peptide-intein with an N-terminal Cys,
as demonstrated by the synthesis of kalata B1.28
Here we describe two amide-to-amide schemes in the
synthesis of cyclic cysteine peptides, a cyclotide hedyotide B1
(hB1) and a sunflower seed trypsin inhibitor SFTI-1. In the
biological scheme using an intein, we investigated the intein
splicing site specificity to optimize its biosynthesis of the
cyclotide hB1. In the chemical scheme, we applied the
thioethylbutylamido (TEBA) as a C-terminal linker for preparing
SFTI-1 and showed that the amide-to-amide cyclization can be
performed in a one-pot reaction using the crude peptide after its
cleavage from the resin support.
The chemical scheme of an amide-to-amide approach to
prepare peptide macrocycles employs an amide bond as a latent
thioester to form a new amide bond, mimicking an intein-based
peptide cyclization scheme (Scheme 1). The use of amides as
latent thioesters is best exemplified by N-alkyl cysteine,
secondary cysteinyl amides that have been successfully exploited
by Hojo and his coworkers in solid-phase peptide synthesis using

3
Hedyotide B1 (hB1), the first bracelet cyclotide identified
ACCEPTED MANUSCRIPT
from herbal plant Hedyotis biflora, is a cyclic cysteine-knot
peptide of 30 residues.43 For an intein-based synthesis of hB1,
each of the six Cys residues can serve as a ligation site for
macrocyclization using a Cys-peptide-intein linear precursor. To
determine the optimal ligation site, we prepared all six linear
precursors of hB1, named as B1X (X = R, T, P, G, Y, and F),
based on the six Xaa-Cys dipeptide sequences in hB1 (Xaa =
Arg, Thr, Pro, Gly, Tyr and Phe), each with an N-terminal Cys
residues and a different Xaa-residue at the C-terminus (Table 1).
These precursors were co-expressed with intein Mxe that linked
to a chitin-binding domain (CBD) in an E. coli system (Scheme
3). The fusion protein B1X-Mxe-CBD was isolated in a chitin
column, where the enzyme-catalyzed thioesterification and
thiolysis took place. The reducing thiol sodium 2-
mercaptoethanesulfonate (MESNa) was added to facilitate B1X
release by thiolysis, a thiol-thioester reaction. It also protected
multiple Cys residues from disulfide formation. Previously we
have shown that cysteine-rich peptides, both the N-terminal and
internal Cys side-chain thiol groups can undergo an
intramolecular thiol-thioester reaction, an S-S acyl shift reaction
with the C-terminal thioester, leading to the formation of multiple
peptide thiolactones.21 These thiolactones facilitate the
cyclization reaction through a thia zip mechanism of thiolactone
ring expansion via a series of entropy-driven thiol-thiolactone
exchange reactions.19 The thiolactone ring expansion leads to an
end-to-end thiolactone with the Nα-amine and Cα-carboxyl
brought into close proximity to allow an S-N acyl shift reaction
to occur resulting in a peptide macrocycle.
Scheme 2. A scheme demonstrating the biomimetic peptide
cyclization mediated by thiolethylamido (TEA) group. Three acyl
shift reactions (N-S, S-S, S-N) mediate the breaking and making of
the peptide bond. Through a series of acyl shifts, the thioester
surrogate was excluded from the target peptide with the N and C
termini joined by an amide bond.
2. Results and Discussion
2.1. Intein-mediated synthesis of hedyotide B1

4
Tetrahedron
ACCEPTED MANUSCRIPT
Table 1. Six hedyotide B1 precursors B1X (X for R, T, P, G, Y, F) with different ligation site residues
Precursor
Amino acid sequences
I
II
III
IV V
VI
hB1
GTR C GET C FVLP C WSAKFG C Y C QKGF C YRN
C GET C FVLP C WSAKFG C Y C QKGF C YRNGTR
C FVLP C WSAKFG C Y C QKGF C YRNGTR C GET
C WSAKFG C Y C QKGF C YRNGTR C GET C FVLP
C Y C QKGF C YRNGTR C GET C FVLP C WSAKFG
C QKGF C YRNGTR C GET C FVLP C WSAKFG C Y
C YRNGTR C GET C FVLP C WSAKFG C Y C QKGF
B1R
B1T
B1P
B1G
B1Y
B1F
Cysteine residues (in yellow) are aligned with the native sequence.
Different C-terminal residues are highlighted in red.
For the release of B1X from the intein, we found that both
inter- and intramolecular thiolysis occurred under neutral to basic
pH. For our experiments, the release of B1X was performed by
an intermolecular thiolysis using a large excess of MESNa and
the amount released at 12 h and 24 h intervals was analyzed by
SDS-PAGE based on the intact fusion protein and intein-CBD
which remained on the chitin-beads (Fig. 1A). Our results
showed that the cleavage of B1G (X = Gly) from fusion protein
was complete within 12 h whereas cleavage of B1R, B1Y or B1F
(X = Arg, Tyr and Phe, respectively) was incomplete in 12 h,
with yields about 50%, but increased to >80% after 24 h (Fig.
1B). The thiolytic release from the fusion protein of both B1T (X
= Thr) and B1P (X = Pro) was slow, with yields about 60% for
B1T and 50% for B1P after 24 h.
Scheme 3. Schematic illustration of Intein-mediated thioester
formation and enzyme-free thia zip cyclization. Cysteine-rich
peptides (N-terminal cysteine in orange and C-terminal residues in
green) were co-expressed with intein (red) and chitin-binding
domain (brown) in E. coli. The fusion proteins were extracted and
purified on the affinity column containing chitin beads (blue). Intein-
mediated N-S acyl shift reaction resulted in formation of peptide-
thioesters that were cleaved from the fusion proteins through S-S
acyl shift reactions either by MESNa-mediated thiolysis that
produced MES-peptide thioester or by internal thiol-mediated
thiolactone formation. The on-column cyclization reaction proceeded
through thia zip mechanism and resulted in end-to-end cyclic
cysteine-rich peptides.
Fig. 1. Comparison of cleavage efficiency of six B1X precursors.
(A) SDS-PAGE analysis of proteins remained on chitin column.
Protein samples were prepared by dissolving 30 ꢀl chitin beads in
15 ꢀl 2x loading buffer and heating at 100 °C for 1 min. The odd-
number lane of each clone shows cleavage ratio after 12 h
incubation in column buffer containing MESNa and the even-
number lane shows the cleavage ration after 24 h incubation. (B)
Comparison of cleavage efficiency of six B1X-Mxe-CBD fusion
proteins.

5
Next, we identified the products released from the thiolytic
ACCEPTED MANUSCRIPT
reactions by HPLC and MS (Fig. 2A). We anticipated that the
B1X-thioesters (MES thioesters) and/or B1X-thiolactones would
undergo the thia zip mediated macrocyclization to form the end-
to-end lactam, the cyclic but S-reduced form of hB1. Indeed,
HPLC analysis showed that the cyclic hB1 was obtained as the
major product for B1G, B1Y, B1F, B1R (>90%) and B1T (>
60%). In contrast, B1P afforded <5% yield of cyclic B1P, with
B1P-OH, the hydrolyzed product (60% yield), as the major side
product together with about 30% yield of the unreacted thioester
B1P-MES (Fig. 2B). By calculating the peptide release from the
intein and cyclization yield, we obtained the ranking of
cyclization efficiency as Gly>Arg, Tyr, Phe>Thr>>>Pro with a
difference of 20-fold between B1G and B1P. Our results suggest
that in designing a Cys-peptide-intein construct, placing a Gly at
the C-terminus as the C-terminal ligation site is highly desirable
while a hindered amino acid such as Thr or Pro undesirable.
Fig. 3. Global oxidative folding of hedyotide B1. Reduced
hedyotide B1 (peak R) was eluted at 42 min and native hedyotide
B1 (peak N) at 43.8 min.
The fully reduced cyclic hB1 was subjected to global
oxidative folding with 10 ꢀM cyclized hB1 in the presence of
reduced/oxidized glutathione (molar ratio peptide:GSH:GSSG =
1:10:100) and 70% isopropanol at pH 8.5, an optimized condition
which we have previously reported.12 The reaction was
completed after 48 h at room temperature to give 48% yield of
the native-folded hB1 (Fig. 3, peak N). Cyclotides contain an
inner sulfur core consisting of a cysteine knot, which displaces
the bulky hydrophobic side chains to its surface. Consequently, a
cyclotide such as hB1 contains hydrophobic patches and tends to
aggregate in the folding process. Addition of co-solvent
isopropanol improved the folding yield in two aspects, increasing
solubility of peptide by reducing non-covalent interactions, and
facilitated thiol-disulfide exchange reactions by promoting the
native-like conformation of hB1. It should be noted that hB1
belongs to the bracelet family of cyclotides. Comparing with
Möbius family of cyclotides, oxidative folding of bracelet
cyclotides has been known to be difficult to obtain high
yields.17,44
2.2. TEBA-mediated synthesis of SFTI-1
Sunflower trypsin inhibitor-1 (SFTI-1) is a 14-residue cyclic
peptide with one disulfide linking Cys3 and Cys11 to stabilize
two β-sheets. It is one of the most potent trypsin inhibitor (Ki=0.1
nM) belonging to the Bowman-Birk family.6 It is relatively stable
against enzyme degradation and could be a useful template for
peptide engineering for therapeutics.
In our synthesis of SFTI-1 utilizing a precursor design of Cys-
peptide-TEBA,
we
started
with
coupling
2-
(butylamino)ethanethiol (TEBA amine form) to the Cl-Trt(2-Cl)
resin to afford the TEBA resin (Scheme 4). 2-
(butylamino)ethanethiol is commercially available and readily
reacted with Cl-Trt(2-Cl) resin 1 to give TEBA-Trt(2-Cl) resin 2
with a thioether linkage in 1 h. An advantage of this approach
over an N-alkyl Cys as a linker is that the resin support serves as
a protecting group (Trt) for the thiol moiety of TEBA, which is
deprotected during the TFA cleavage step concurrent with other
protecting groups of the peptide. In the Cys-peptide-TEBA
scheme, there are two logical ligation sites, Cys-Arg and Cys-Ile.
Based on the lessons learned from the cyclotide hB1 synthesis
using intein, Arg was chosen over the β-branched amino acid Ile
as the ligation site to minimize steric hindrance in the cyclization
step.
Fig. 2. Comparison of cyclization efficiency of six B1X
precursors. (A) HPLC profiles of product mixtures eluted from
chitin column. (B) Cyclization products analysis and yield.

6
Tetrahedron
containing all three products 5, 6a and 6b were subjected to the
ACCEPTED MANUSCRIPT
thia zip cyclization at pH 7 at room temperature to afford 7 after
4 h. After HPLC purification and oxidative folding with 20%
DMSO, native SFTI-1 8 was obtained in 33% yield.
With two cysteine residues in the SFTI-1 sequence, we
anticipated that they could undergo a direct thiolactone formation
to release the TEBA linker, rendering the use of an external thiol
for the extra step of thioesterification unnecessary. Thus, the
second approach in the macrocyclization of Cys-SFTI-TEBA
was performed without MESNa or any external thiol under the
experimental conditions similar as the previous approach. After
15 h, except for the thioester 5, all expected products, SFTI-
TEBA thioester 4b, thiolactones 6a and 6b, were found. A small
amount of hydrolysis product (<5%) eluted earlier than
thiolactones 6a in the HPLC trace was also observed. After
cyclization at pH 7, the reaction mixture containing product 7
was directly subjected to oxidative folding with 20% DMSO
without purification. Both approach with or without the use of an
external thiol such as MESNa gave comparable cyclization yield,
which was 89% and 84%, respectively. However, under our
experimental conditions, the approach without an external thiol
required only one purification step from the linear precursor to
native product whereas the other approach required two
purification steps. Thus the overall yield was 10% higher (43%
vs 33% isolated yield) when no external thiol was included in the
reaction.
Scheme 4. A scheme of the TEBA-mediated synthesis of cyclic
peptide SFTI-1.
The synthesis of Cys-SFTI-TEBA commenced with the
coupling of Fmoc-Arg(Pbf)-OH as the first amino acid to the
TEBA resin using HATU as the coupling reagent. The reaction
completed in 1 h as determined by the chloranil test, which
detects the presence of free secondary amine. The Fmoc-
Arg(Pbf)-TEBA resin 3 was then used as the starting material for
the synthesis of SFTI-1 in a microwave peptide synthesizer by a
Fmoc-chemistry protocol with a modified deprotection step using
20% morpholine instead of 20% piperidine (see experimental
section 4.8). SFTI-1 contains a base-sensitive Asp-Gly sequence,
which can lead to aspartimide formation during synthesis. To
suppress the aspartimide formation, we employed a milder base
in the deprotection steps, using morpholine instead of piperidine.
In this case, less than 1% aspartimide formation was observed.
After the completion of the SFTI-1 synthesis and its release from
the resin support by TFA, HPLC analysis showed that two
products, the amide form (N form) SFTI-TEBA 4a and thioester
form (S form) 4b, were obtained.
The cleaved peptides were subjected to cyclization in two
approaches (Scheme 4). The first approach was a two-step
process, a process similar to intein approach with the thiolytic
release of peptide as a thioester by a large excess of an external
thiol, followed by cyclization. Indeed, the Cys-SFTI-TEBA was
released from the TEBA linker as Cys-SFTI-MES thioester using
a large excess of MESNa. The thiolytic release was achieved at
40 °C, pH 3.22 10 mM TCEP was added to prevent disulfide bond
formation (Fig. 4). The thioester SFTI-MES 5 was observed as
the major product after 15 h. Subsequently, two thiolactones (6a
and 6b) were also generated through thiol-thioester exchange
between the thiol side chain of Cys residues and the C-terminal
thioester bond, a 19-atom thiolactone 6a generated from internal
Cys and a 43-atom thiolactone 6b generated from the N-terminal
Cys. The thiolactones and the end-to-end lactam SFTI-1 7 (the
reduced cyclic form) are isomers with the same MW 1515 Da,
but thiolactones can be distinguished from lactam by the stronger
UV-absorbance at 260 nm, a characteristic absorbance of a
thioester bond. In HPLC trace, product 6b eluted close to 7
whereas 6a eluted earlier than 6b. The product mixtures
Fig. 4. RP-HPLC profiles of synthetic intermediates in the
synthesis of SFTI-1. Starting material was subjected to two
conditions with or without MESNa at 40 °C pH 3 for 18 h. After
the pH adjusted to 7, the cyclization was conducted at room
temperature for 4 h. Then 20% DMSO was used to perform the
oxidative folding reaction. Each intermediates were monitored by
RP-HPLC.

7
3. Conclusions
integrated into pTXB1 by NdeI/SapI double digestion and T4
ACCEPTED MANUSCRIPT
DNA ligation. Ligation products were transformed into
competent DH5α E. coli. Correct clones were screened by
EcoRI/EcoRV digestion (1 h, 37 °C) and agarose-electrophoresis
(Fig. S3). The purified plasmids were transformed into high-
performance expression strain ER2566 (BL21).
Our chemical scheme of an amide-to-amide approach in
peptide macrocyclization has two major features in common with
the biological approach using an intein. First, the design of the
linear precursor construct of Cys-peptide-intein/TEBA is similar,
although the size of TEBA is several hundred fold smaller than
an intein. Second, they share similar chemical principle in the
amide-to-amide splicing process via a series of acyl shifts,
resulting in the excision of the C-terminal amide moiety as an
intein or the TEBA linker, followed by forming a new amide
cysteinyl bond to generate a peptide macrocycle.
4.3. IPTG-induced expression and fusion protein extraction
One fresh-growing colony was innoculated into 3 ml LB
medium in a 15 ml culture tube for 3 h at 37 °C. One ml culture
was then transferred to 500 ml LB medium and shaking-
incubated in a 2 L flask. Until A600 reached 0.7, decreased
incubator temperature to 16 °C and added 0.3 mM IPTG, shaked
for 24 h at 150 rpm.
There are also significant differences between these two
methods. Whereas an intein catalyzes the amide to thioester at the
C-terminal junction at neutral pH, the TEBA linker requires
fairly acidic conditions for transforming an amide to a thioester
via an N-S acyl shift. The N-terminal cysteinyl thiol, and in
cysteine-rich peptides, an internally placed cysteinyl thiol, play
an important role to release or excise either an intein or TEBA
linker to complete the excision process through one or more
thiol-thioester exchange reactions. Thus, the amide-to-amide
chemical approach mediated by the TEBA is particular suitable
for preparing cyclic cysteine peptides since the presence of
multiple internal thiols greatly facilitates the release the TEBA
linker as thiolactones which can then directly undergo thia zip
cyclization. Again, there is a large pH difference between the
chemical means to release the TEBA and biological means to
release the intein, with the former at an acidic pH and the later in
neutral pH. Thus, the TEBA-mediated scheme depends on the
catalysis by general acids and bases and differs from catalysis by
specific acid and base of intein to break and make peptide bonds
in peptide macrocyclization.
Bacterial cells were collected by 10 min centrifugation at
10,000 rcf. The pellets were resuspended in cell lysis buffer (20
mM Tris-Cl, pH 8.5, 500 mM NaCl, 1 mM EDTA, 0.1% Triton
X-100) and lysed by ultrasonication for 1 h on ice (4s-
operation/10s-pause). The clarified cell extract was purified by
40 min centrifugation of the cell lysate at 10,000 rcf at 4 °C and
filtration (0.22 µm). Expressed proteins in the cell lysates and
clarified extracts were examined by SDS-PAGE (Fig. S4).
4.4. On-column purification and cleavage
Chitin beads (10 ml) were loaded into chitin-binding column
and then calibrated by 100 ml column buffer (20 mM Tris-Cl,
500 mM NaCl, 1 mM EDTA, pH 8.5). Pre-chilled 50ml clarified
cell extract from 1 L cell culture was slowly loaded on to the
chitin column (flow rate <0.5 ml/min). Chitin beads were washed
with at least 200 ml column buffer (flow rate = 2 ml/min) to
stabilize the affinity binding between chitin and CBD. After
wash, beads were flushed with 50 ml column buffer to remove
unspecific binding. This purification step was conducted under 4
°C in the cold room. Chitin beads were then flushed with 20 ml
cleavage buffer containing mercaptoethanesulfonate (20 mM
Tris-Cl, 500 mM NaCl, 1 mM MESNa, pH 8.5) and incubated at
room temperature for 24 h, or at 4 °C for not more than 48 h. The
cleavage efficiency was checked by analyzing the column-bound
proteins using SDS-PAGE. Splicing products were eluted with
10 ml cleavage buffer. The eluents were desalted using SPE
ODS-C18 Cartridges (1 ml) and eluted with 1 ml 60% ACN,
0.05% TFA, pH 2.0. TCEP (30 mM) was added to the eluents
and incubated for 1 h at 37 °C to reduce any unexpected disulfide
bonds.
4. Experimental section
4.1. General
All Fmoc-amino acids reagents and solvents were used
without purification. Amino acids were purchased from Chem-
Impex. HOBt, PyBOP and HATU were purchased from GL
Biochem. Chlorotrityl chloride resin (1.2 mmol/g) and solvents
DMF, DCM, piperidine, morpholine, and diethylether were
purchased from Merck. All HPLC was performed on Shimadzu
liquid chromatography using an analytical column (4.6 mm x 250
mm) with a linear gradient of 10 to 60% acetonitrile (ACN)/0.1%
trifluoroacetic acid (TFA) for 25 min at a flow rate of 1 ml/min,
and a preparative column C18 column (22 mm x 250 mm) at a
flow rate of 6 ml/min. Molecular weights were measured with
4800 plus MALDI-TOF/TOFTM Analyzer.
4.5. Tricine SDS-PAGE
Samples for SDS-PAGE included cell pellet from 1 ml cell
culture before induction, cell pellet from 0.5 ml cell culture after
induction, cell debris from 50 ꢀl crude extract, clarified extract
from 50 ꢀl crude extract, 20 ꢀl chitin beads collected after step
sample loading, and 20 ꢀl chitin beads after cleavage. Six
samples were resuspended/mixed with 40 ꢀl 2x sample loading
buffer (Biorad, tricine-SDS sample buffer) and heated at 100 °C
for 10 min. Tricine-SDS separation gel contained 10%
acrylamide/bis-acrylamide mix (29:1 stock solution), 1M Tris-
Cl/0.1% SDS (w/v, 3x stock solution, pH 8.5), 20% glycerol,
0.05% ammonium persulfate (APS) (w/v, 10% stock solution),
and 0.005% TEMED (w/v). Tricine-SDS stacking gel contained
4.05% acrylamide/bis-acrylamide mix (29:1 stock solution),
7.5% Tris-Cl/0.1% SDS (w/v, 3x stock slution, pH 8.5), 0.05%
ammonium persulfate (APS, w/v, 10% stock solution), and 0.01%
TEMED. The 5x Tricine-SDS PAGE running buffer stock was
prepared by 0.5M tricine, 0.5M Tris-base, 0.5% SDS (w/v).
4.2. Cloning of Hedyotide B1 precursors
The sequence of B1 cDNA (cB1) has been confirmed by 3’-
RACE and 5’-RACE reactions from previous study. The cDNA
sequence was optimized according to E. coli codon bias index
(Table S1) by modifications on G1, R3, P12 and R29. The six
hB1 precursors B1X (X=R, T, P, G, Y, and F) were synthesized
by tandem-extension PCR without adding templates in the PCR
mix (Fig. S1). For each precursor, four primers were designed to
fully cover the complete nucleotide sequence with overlapping
oligos. Primers 1f and 4r carries the NdeI and SapI restriction
sites respectively for integrating into vector pTXB1 (NEB,
IMPACT kit) (Fig. S2). PCR products were purified after 25
cycles of amplification by agarose-gel (2% w/v) DNA extraction
(Invitrogen, PureLink™ Quick Gel Extraction Kit). The PCR
products obtained from each tandem-extension reaction were

8
Tetrahedron
4.6. RP-HPLC analysis
added directly to the cyclization mixture to perform the oxidative
ACCEPTED MANUSCRIPT
folding reaction. For both conditions, the oxidative folding
reaction was quenched with 1 N HCl to pH 2 and purified by
using RP-HPLC to give native SFTI-1.
Analytical RP-HPLC was performed using C18 column (5
µm, 300 Ǻ, 4.6x250 mm) with a flow rate of 1 ml/min. The
mobile phase included buffer A (0.05% TFA, milli-Q water) and
buffer B (80% ACN, 0.035% TFA, milli-Q water). The gradient
for cyclic products was 10-70% ACN for 60 min. The gradient
for folding products was 0-50% ACN for 50 min.
Acknowledgments
This research was supported in part by Biomedical Research
Council (BMRC 09/1/22/19/612) of A*STAR and National
Research Foundation (CRP8-2011-05) of Singapore.
4.7. Disulfide folding and characterization
Lyophilized B1 precursors were oxidized by GSH and GSSG
in the molar ratio of peptide:GSSG:GSH = 1:10:100. The powder
mixture was dissolved in 70% iPrOH with 0.1 mM Tris-Cl (pH
8.5) and mixed by vortex. The reaction solution was degassed by
nitrogen to eliminate oxygen interference to the thiol-thioester
exhange reaction. Folding reactions were conducted for 48 h at
room temperature.
4.8. Synthesis of SFTI-1 on TEBA-loaded resin
Cl-Trt(2-Cl) resin (417 mg, 0.5 mmol) was swollen for 30 min
in CH2Cl2 (15 ml) and washed with CH2Cl2 (10 ml x 2). The
resin was re-suspended in 10 ml CH2Cl2. To the suspension, 2-
(butylamino)ethanethiol (37.6 µl, 0.25 mmol) was added, and the
suspension was shaken for 1 h at room temperature. Then DIEA
(133 µl, 0.75 mmol) in MeOH (10 ml) was added to the
suspension and the suspension was shaken for 10 min. After
filtration, the resin was washed with CH2Cl2 (10 ml x 3) and
DMF (10 ml x 3). To the resin, a pre-mixed solution containing 4
equiv Fmoc-Arg(Pbf)-OH (M.W. 648.8, 1 mmol 649 mg), 4
equiv HATU (M. W. 380.2, 1 mmol, 380 mg) and 6 equiv DIEA
(265 µl, 0.75 mmol) was added. The coupling reaction was
performed at room temperature for 1 h and monitored by
chloranil test (2% acetaldehyde, 2% chloranil in DMF, 5 min at
room temperature). Peptide elongation was carried out in an
automatic microwave peptide synthesizer using PyBOP/DIEA.
Fmoc deprotection was performed by the use of 20% morpholine
in DMF. 940 mg SFTI-TEBA-loaded resin was produced and
cleaved with 10 ml cleavage cocktail containing TFA/TIS/H2O
(95/2.5/2.5, v/v) to remove the solid support and side chain
protection groups. The cleavage reaction was allowed to proceed
for 2 h and cold diethyl ether was used to precipitate the peptide.
Then it was dried in vacuo and analyzed by RP-HPLC. 330 mg
precipitation was generated and used directly for the following
cyclization reaction.
4.9. Cyclization and folding of SFTI-1
4.9 mg crude peptide precipitation containing SFTI-TEBA 4a
and 4b was dissolved in 0.1M sodium phosphate buffer (pH 3, 1
ml) at a concentration of 3 mM. The reaction mixture was
incubated at 40 °C with or without MESNa (150 µmol, 24.7 mg)
for 18 h. TCEP (10 µmol, 2.9 mg) was added to prevent the
disulfide formation through inter- or intra-molecular linkage. The
formation of synthetic intermediates such as thioesters and
thiolactones was monitored by RP-HPLC at both 220 and 260
nm. Then the reaction mixture was diluted with 0.1M sodium
phosphate (pH 3, 2 ml) and adjusted with 1 N NaOH. After the
pH was adjusted to 7, the thia zip cyclization was allowed to
proceed for 4 h at room temperature. The cyclization process was
monitored by RP-HPLC. After the cyclization completed, the
reaction mixture was purified by RP-HPLC when MESNa was
included as an external thiol. Reduced cyclic SFTI-1 was then
dissolved in pH 7.5 phosphate buffer at concentration of 1 mM
and 20% DMSO (v/v) were added. The oxidative folding reaction
was allowed to proceed at room temperature and monitored by
RP-HPLC. For conditions without MESNa, 20% DMSO was

9
21.
Hemu, X.; Taichi, M.; Qiu, Y.; Liu, D. X.; Tam, J. P.
Pept. Sci. 2013, 100, 492-501.
References and notes
ACCEPTED MANUSCRIPT
1.
Eisenbrandt, R.; Kalkum, M.; Lai, E.-M.; Lurz, R.;
Kado, C. I.; Lanka, E. J. Biol. Chem. 1999, 274, 22548-
22555.
22.
Taichi, M.; Hemu, X.; Qiu, Y.; Tam, J. P. Org. Lett.
2013, 15, 2620-2623.
23.
24.
Haase, J.; Lanka, E. J. Bacteriol. 1997, 179, 5728-5735.
2.
González, C.; Langdon, G. M.; Bruix, M.; Gálvez, A.;
Valdivia, E.; Maqueda, M.; Rico, M. Proc. Natl. Acad.
Sci. U. S. A. 2000, 97, 11221-11226.
Lee, J.; McIntosh, J.; Hathaway, B. J.; Schmidt, E. W.
J. Am. Chem. Soc. 2009, 131, 2122-2124.
3.
4.
5.
6.
Tran, D.; Tran, P. A.; Tang, Y. Q.; Yuan, J.; Cole, T.;
Selsted, M. E. J. Biol. Chem. 2002, 277, 3079-3084.
25.
26.
Barber, C. J. S.; Pujara, P. T.; Reed, D. W.; Chiwocha,
S.; Zhang, H.; Covello, P. S. J. Biol. Chem. 2013, 288,
12500-12510.
Trabi, M.; Craik, D. J. Trends Biochem. Sci. 2002, 27,
132-138.
Saska, I.; Gillon, A. D.; Hatsugai, N.; Dietzgen, R. G.;
Hara-Nishimura, I.; Anderson, M. A.; Craik, D. J. J.
Biol. Chem. 2007, 282, 29721-29728.
Pelegrini, P.; Quirino, B.; Franco, O. Peptides 2007, 28,
1475-1481.
27.
28.
29.
30.
Camarero, J. A.; Muir, T. W. J. Am. Chem. Soc. 1999,
121, 5597-5598.
Luckett, S.; Santiago Garcia, R.; Barker, J. J.;
Konarev, A. V.; Shewry, P. R.; Clarke, A. R.; Brady, R.
L. J. Mol. Biol. 1999, 290, 525-533.
Kimura, R. H.; Tran, A.-T.; Camarero, J. A. Angew.
Chem. Int. Ed. 2006, 45, 973-976.
7.
8.
Hruby, V. J. Life Sci. 1982, 31, 189-199.
Wu, Z.; Guo, X.; Guo, Z. Chem. Commun. 2011, 47,
9218-9220.
Colgrave, M. L.; Craik, D. J. Biochemistry 2004, 43,
5965-5975.
Jia, X.; Kwon, S.; Wang, C. I.; Huang, Y. H.; Chan, L.
Y.; Tan, C. C.; Rosengren, K. J.; Mulvenna, J. P.;
Schroeder, C. I.; Craik, D. J. J. Biol. Chem. 2014, 289,
6627-6638.
9.
Thorstholm, L.; Craik, D. J. Drug Discov. Today Tech.
2012, 9, e13-e21.
10.
11.
Gran, L. Acta Pharmacol. Toxicol. 1973, 33, 400-408.
31.
Hirata, R.; Ohsumk, Y.; Nakano, A.; Kawasaki, H.;
Suzuki, K.; Anraku, Y. J. Biol. Chem. 1990, 265, 6726-
6733.
Cascales, L.; Henriques, S. T.; Kerr, M. C.; Huang, Y.
H.; Sweet, M. J.; Daly, N. L.; Craik, D. J. J. Biol. Chem.
2011, 286, 36932-36943.
32.
33.
Kane, P.; Yamashiro, C.; Wolczyk, D.; Neff, N.; Goebl,
M.; Stevens, T. Science 1990, 250, 651-657.
12.
13.
Wong, C. T. T.; Rowlands, D. K.; Wong, C.-H.; Lo, T.
W. C.; Nguyen, G. K. T.; Li, H.-Y.; Tam, J. P. Angew.
Chem. Int. Ed. 2012, 51, 5620-5624.
Perler, F. B.; Davis, E. O.; Dean, G. E.; Gimble, F. S.;
Jack, W. E.; Neff, N.; Noren, C. J.; Thorner, J.; Belfort,
M. Nucleic Acids Res. 1994, 22, 1125-1127.
Clark, R. J.; Jensen, J.; Nevin, S. T.; Callaghan, B. P.;
Adams, D. J.; Craik, D. J. Angew. Chem. Int. Ed. 2010,
49, 6545-6548.
34.
Xu, M. Q.; Southworth, M. W.; Mersha, F. B.;
Hornstra, L. J. Cell 1993, 75, 1371-1377.
14.
15.
Nguyen, G. K. T.; Lim, W. H.; Nguyen, P. Q. T.; Tam,
J. P. J. Biol. Chem. 2012, 287, 17598-17607.
35.
36.
Xu, M. Q.; Perler, F. B. EMBO J. 1996, 15, 5146-5153.
Nguyen, G. K. T.; Lian, Y.; Pang, E. W.; Nguyen, P. Q.
T.; Tran, T. D.; Tam, J. P. J. Biol. Chem. 2013, 288,
3370-3380.
Dawson, P. E.; Muir, T. W.; Clark-lewis, I.; Kent, S. B.
H. Science 1994, 266, 776-779.
37.
38.
39.
40.
Tam, J. P.; Lu, Y. A.; Liu, C.-F.; Shao, J. Proc. Natl.
Acad. Sci. U. S. A. 1995, 92, 12485-12489.
16.
Tam, J. P.; Lu, Y. A. Tetrahedron Lett. 1997, 38, 5599-
5602.
Muir, T. W.; Sondhi, D.; Cole, P. A. Proc. Natl. Acad.
Sci. U. S. A. 1998, 95, 6705-6710.
17.
18.
Tam, J. P.; Lu, Y. A. Protein Sci. 1998, 7, 1583-1592.
Tam, J. P.; Lu, Y. A.; Yang, J. L.; Chiu, K. W. Proc.
Natl. Acad. Sci. U. S. A. 1999, 96, 8913-8918.
Hojo, H.; Onuma, Y.; Akimoto, Y.; Nakahara, Y.;
Nakahara, Y. Tetrahedron Lett. 2007, 48, 25-28.
19.
20.
Tam, J. P.; Lu, Y. A.; Yu, Q. J. Am. Chem. Soc. 1999,
121, 4316-4324.
Nakamura, K.; Sumida, M.; Kawakami, T.; Vorherr,
T.; Aimoto, S. Bull. Chem. Soc. Jpn. 2006, 79, 1773-
1780.
Tam, J. P.; Wong, C. T. T. J. Biol. Chem. 2012, 287,
27020-27025.
41.
Ollivier, N.; Dheur, J.; Mhidia, R.; Blanpain, A.;
Melnyk, O. Org. Lett. 2010, 12, 5238-5241.

10
Tetrahedron
42. Hou, W.; Zhang, X.; Li, F.; Liu, C.-F. Org. Lett. 2011,
These data include experimental procedures, characterization
ACCEPTED MANUSCRIPT
13, 386-389.
of PCR products and plasmids by agarose gel, SDS-PAGE gel
analysis of expressed fusion proteins, all MALDI-TOF mass
spectra of reduced-cyclic/hydrolyzed/native hB1 and B1-MES
thioester, and chemically synthesized compounds 4-8.
43.
44.
Wong, C. T. T.; Taichi, M.; Nishio, H.; Nishiuchi, Y.;
Tam, J. P. Biochemistry 2011, 50, 7275-7283.
Gunasekera, S.; Daly, N. L.; Clark, R. J.; Craik, D. J.
Antioxid. Redox Signal 2009, 11, 971-980.
Supplementary Material

ACCEPTED MANUSCRIPT
Supplementary material
Peptide macrocyclization through transpeptidation
Xinya Hemua , Yibo Qiua, and James P. Tama *
aSchool of Biological Sciences, Nanyang Technological University, 60 Nanyang Drive, 03s-71,
Singapore 637551
Table of Content
Table S1. Primer design for tandem-extension PCR to amplify CRP sequences flanked by NdeI/SapI
restriction sites
Figure S1. Schema illustration of tandem-extension PCR
Figure S2. Preparation of recombinant plasmid pTXB1-B1x
Figure S3. EcoRI/EcoRV double digestion screening of plasmids with PCR product insertion
Figure S4. Expression and extraction of fusion protein B1x-Mxe-CBD
Figure S5. MALDI-TOF spectra of reduced cyclic hB1, hydrolyzed hB1x, B1P-MES thioester and native-
folded cyclic hB1
Figure S6. MALDI-TOF spectra of synthetic compound 4-8

ACCEPTED MANUSCRIPT
Table S1. Primer design for tandem-extension PCR to amplify CRP sequences flanked by
NdeI/SapI restriction sites. The longer nucleotide lengths and higher GC content of the
complementary region on 2f/3r conferred higher melting temperature (Tm) for the first extension
reaction to prepare double-strand template for the second extension. The annealing temperatures based
on Tm were set as 54˚C for 1st reaction and 47˚C for the 2nd reaction.
Primer
Sequence
Tm (˚C)
B1R-1f
B1R-2f
B1R-3r
B1R-4r
5’-GGTGGTCATATGTGCGGTGAGACTTGCTTCGTTTTACCGTG
5’-TGCTTCGTTTTACCGTGCTGGTCCGCCAAGTTCGGCTGCTACTGCC
5’-CCGTTGCGGTAACAAAAACCCTTTTGGCAGTAGCAGCCGAACTT
5’-GGTGGTTGCTCTTCCGCAGCGAGTGCCGTTGCGGTAACAAA
51.3
59.1
53.1
B1T-1f
B1T-2f
B1T-3r
B1T-4r
5’-GGTGGTCATATGTGCTTCGTTTTACCGTGCTGGTCCGCCAAGTTCGG
5’-TCCGCCAAGTTCGGCTGCTACTGCCAAAAGGGTTTTTGTTAC
5’-ACCGCAGCGAGTGCCGTTGCGGTAACAAAAACCCTTTTGGCA
5’-GGTGGTTGCTCTTCCGCAAGTCTCACCGCAGCGAGTGC
53.8
57.6
54.1
5’-GGTGGTCATATGTGCTGGTCCGCCAAGTTCGGCTGCTACTGCC
5’-TCGGCTGCTACTGCCAAAAGGGTTTTTGTTACCGCAACGGCAC
5’-AACGAAGCAAGTCTCACCGCAGCGAGTGCCGTTGCGGTAACA
5’-GGTGGTTGCTCTTCCGCACGGTAAAACGAAGCAAGTCTCACC
B1P-1f
B1P-2f
B1P-3r
B1P-4r
53.2
58.6
51.9
5’-GGTGGTCATATGTGCTACTGCCAAAAGGGTTTTTGTTACCGCAACGG
5’-TTTTGTTACCGCAACGGCACTCGCTGCGGTGAGACTTGCTTCG
5’-GAACTTGGCGGACCAGCACGGTAAAACGAAGCAAGTCTCACCGC
5’-GGTGGTTGCTCTTCCGCAGCCGAACTTGGCGGACCA
B1G-1f
B1G-2f
B1G-3r
B1G-4r
53.1
59.1
52.5
5’-GGTGGTCATATGTGCCAAAAGGGTTTTTGTTACCGCAACGG
5’-TTTGTTACCGCAACGGCACTCGCTGCGGTGAGACTTGCTTCG
5’-GAACTTGGCGGACCAGCACGGTAAAACGAAGCAAGTCTCACCGC
5’-GGTGGTTGCTCTTCCGCAGTAGCAGCCGAACTTGGCGGACCA
B1Y-1f
B1Y-2f
B1Y-3r
B1Y-4r
53.1
59.1
52.5
B1F-1f
B1F-2f
B1F-3r
B1F-4r
5’-GGTGGTCATATGTGTTACCGCAACGGCACTCGCTGCGGTGAG
5’-ACTCGCTGCGGTGAGACTTGCTTCGTTTTACCGTGCTGGTCCGC
5’-TTTGGCAGTAGCAGCCGAACTTGGCGGACCAGCACGGT
5’-GGTGGTTGCTCTTCCGCAAAAACCCTTTTGGCAGTAGCAGCC
53.0
58.6
54.3
NdeI/SapI restriction sequences on primers 1f/4r were in bold. Complementary sequences on 2f/3r
pair for the first annealing were highlighted in blue. The annealing sequences for the second extension
reaction on 1f/2f and 3r/4r were highlighted with red and orange, respectively.

ACCEPTED MANUSCRIPT
Figure S1. Schema illustration of tandem-extension PCR. Forward primer p1f carried NdeI
restriction site and 5’-end of CRP cDNA. Forward primer p2f contained 14-17bp overlapping
sequence that complementary to p1f (red) and 15-21bp overlapping sequence that complementary to
p3r (blue). Reversed primer p3r overlapped with p2f (blue) and p4r (orange). Reversed primer p4r
overlapped with p3r (orgnge) and carried SapI restriction site on its 3’-end. In the first chain reaction
with higher annealing temperature (56-58°C), only p2f and p3r participated in the polymerization to
generate amplicon 1 for 5 runs. In the second chain reaction, p1f and p4r annealed with amplicon 1 at
lower annealing temperature (51-53°C) to produce amplicon 2 and 3. In the following runs, all four
primers participated in extension reactions to produce amplicon 4, the expected product that possessed
the full length of CRP cDNA sequences flanked by NdeI/SapI restriction sites.

ACCEPTED MANUSCRIPT
Figure S2. Preparation of recombinant plasmid pTXB1-B1x. The PCR products B1x that flanked
with enzyme binding sites (orange) and pTXB1 vector (MCS in blue, Mxe intein in red) were
subjected to NdeI/SapI double digestion. Ligation of the digested products gave a recombinant
plasmid with the in-frame insertion of B1x sequence upstream of Mxe intein.

ACCEPTED MANUSCRIPT
Figure S3. EcoRI/EcoRV double digestion screening of plasmids with PCR product insertion.
Vector pTXB1 was used as negative control, which was digested into two fragments. Five colonies
were picked for overnight culture. Plasmids extracted by mini-prep were subject to double digestion
and agarose gel electrophoresis. Plasmids 1, 4 and 5 were suggested to contain PCR product insertion,
which replaced MCS region containing EcoRI restriction site. Plasmids 2 and 3 were empty vectors.

ACCEPTED MANUSCRIPT
A
B
Figure S4. Expression and extraction of fusion protein B1X-Mxe-CBD. (A) The overall expression
level of fusion proteins (X for R, T, P, G, Y or F). Lanes in odd number: cell lysate before IPTG
induction. Lanes in even number: cell lysate after IPTG induction (B) The solubility of single-intein
fusion protein B1R-Mxe CBD. Lane 1: cell lysate before induction. Lane 2: cell lysate after induction.
Lane 3: insoluble cell pellet containing inclusion bodies. Lane 4: clarified cell extraction.

ACCEPTED MANUSCRIPT
Cyclic reduced hB1 (Calc. 3410.8, Fd. 3410.3)
Linear hydrolyzed hB1 (Calc. 3428.9, fd 3428.1)

ACCEPTED MANUSCRIPT
B1P-MES thioester (Calc. 3551.2, Fd. 3552.1)
Native-folded cyclic hB1 (Calc. 3403.2, Fd. 3403.5)
Figure S5. MALDI-TOF spectra of cyclic reduced hB1, hydrolyzed reduced hB1x, B1P-MES
thioester and native-folded cyclic hB1.

ACCEPTED MANUSCRIPT
4a (calc. 1647.8, Fd. 1649.0)
4b (calc. 1647.8, fd. 1649.0)

ACCEPTED MANUSCRIPT
5 (calc. 1656.8, fd. 1658.0)
6a (calc. 1514.8, fd. 1515.9)

ACCEPTED MANUSCRIPT
6b (calc. 1514.8, fd. 1515.9)
7 (calc. 1514.8, fd. 1515.8)

ACCEPTED MANUSCRIPT
8 (calc. 1512.8, fd. 1513.7)
Figure S6. MALDI-TOF spectra of SFTI-TEBA (N-form 4a), SFTI-TEBA (S-form 4b), SFTI-MES
thioester (5), 19-atom SFTI-thiolactone (6a), 43-atom SFTI-thiolactone (6b), cyclic reduced SFTI (7)
and cyclic folded SFTI (8).