Peptiligase, an Enzyme for Efficient Chemoenzymatic Peptide Synthesis and Cyclization in Water

Article abstract of DOI:10.1002/adsc.201600017

We describe a novel, organic cosolvent-stable and cation-independent engineered enzyme for peptide coupling reactions. The enzyme is a variant of a stable calcium-independent mutant of subtilisin BPN′, with the catalytic Ser212 mutated to Cys and Pro216 converted to Ala. The enzyme, called peptiligase, catalyzes exceptionally efficient peptide coupling in water with a surprisingly high synthesis over hydrolysis (S/H) ratio. The S/H ratio of the peptide ligation reaction is correlated to the length of the peptide substrate and proved to be >100 for the synthesis of a 13-mer peptide, which corresponds to >99% conversion to the ligated peptide product and <1% hydrolytic side-reaction. Furthermore, peptiligase does not require a particular recognition motif resulting in a broadly applicable and traceless peptide ligation technology. Peptiligase is very robust, easy to produce in Bacillus subtilis, and its purification is straightforward. It shows good activity and stability in the presence of organic cosolvents and chelating or denaturing agents, enabling the ligation of poorly soluble (hydrophobic) or folded peptides. This enzyme could be useful for the (industrial) synthesis of diverse (pharmaceutical) peptides. In addition, peptiligase is able to efficiently catalyze head-to-tail peptide cyclization reactions. (Figure presented.) .

Full text of DOI:10.1002/adsc.201600017

FULL PAPERS
DOI: 10.1002/adsc.201600017
Peptiligase, an Enzyme for Efficient Chemoenzymatic Peptide
Synthesis and Cyclization in Water
a
b
b
a,c,
Ana Toplak, Timo Nuijens, Peter J. L. M. Quaedflieg, Bian Wu, *
a,
and Dick B. Janssen *
Biotransformation and Biocatalysis, Groningen Biomolecular Science and Biotechnology Institute, University of
a
Groningen, Nijenborgh 4, 9747 AG Groningen, The Netherlands
Fax : (+31)-50-363-4165; phone: (+31)-50-363-4008; e-mail: d.b.janssen@rug.nl
EnzyPep B. V., Brightlands Campus,Urmonderbaan 22, 6167 RD Geleen, The Netherlands
Present address: CAS Key Laboratory of Microbial Physiology and Metabolic Engineering, Institute of Microbiology,
b
c
Chinese Academy of Sciences, China State Key Laboratory of Transduction Technology, Chinese Academy of Sciences,
Beijing 100101, Peopleꢀs Republic of China
Fax : (+86)-10-6480-7468; phone: (+86)-10-6480-6035; e-mail: wub@im.ac.cn
Received: January 6, 2016; Revised: April 5, 2016; Published online: && &&, 0000
Supporting information for this article is available on the WWW under http://dx.doi.org/10.1002/adsc.201600017.
Abstract: We describe a novel, organic cosolvent- resulting in a broadly applicable and traceless pep-
stable and cation-independent engineered enzyme tide ligation technology. Peptiligase is very robust,
for peptide coupling reactions. The enzyme is a var- easy to produce in Bacillus subtilis, and its purifica-
iant of a stable calcium-independent mutant of subti- tion is straightforward. It shows good activity and
lisin BPN’, with the catalytic Ser212 mutated to Cys stability in the presence of organic cosolvents and
and Pro216 converted to Ala. The enzyme, called chelating or denaturing agents, enabling the ligation
peptiligase, catalyzes exceptionally efficient peptide of poorly soluble (hydrophobic) or folded peptides.
coupling in water with a surprisingly high synthesis This enzyme could be useful for the (industrial) syn-
over hydrolysis (S/H) ratio. The S/H ratio of the pep- thesis of diverse (pharmaceutical) peptides. In addi-
tide ligation reaction is correlated to the length of tion, peptiligase is able to efficiently catalyze head-
the peptide substrate and proved to be >100 for the to-tail peptide cyclization reactions.
synthesis of a 13-mer peptide, which corresponds to
>
99% conversion to the ligated peptide product and
<
1% hydrolytic side-reaction. Furthermore, peptili- Keywords: calcium-free peptide ligase; chemoenzy-
gase does not require a particular recognition motif matic peptide synthesis; peptide cyclization
Introduction
quires fully protected peptides, which are difficult to
dissolve or purify due to their hydrophobicity. Protec-
There is a large and growing interest in pharmaceuti- tion and deprotection steps also reduce the overall
cal peptides over the past decades, with about 60 ap- synthetic yield and require the use of auxiliary chemi-
proved peptide drugs currently on the market and cals, reducing process economy.
some 140 candidates in different phases of clinical de-
Chemoenzymatic peptide synthesis (CEPS), where-
velopment. By 2018, the worldwide peptide drugs in chemically synthesized peptide segments are ligat-
[
1]
market is expected to reach $18 billion. The impor- ed enzymatically, is potentially one of the most cost-
tance of peptides is also recognized in the nutritional efficient technologies for the synthesis of medium-
[
2]
[4]
and cosmetic industries. Despite the high demand sized and long peptides (20–50 amino acids). As
for peptides, their large-scale manufacturing remains compared to chemical segment coupling, CEPS offers
challenging.
[3]
several notable advantages, i.e., protection of the
Complete solid phase or solution phase synthesis of side-chain functionalities is not necessary, racemiza-
long peptides usually leads to accumulation of by- tion is absent, and it is more environmentally friendly.
products and cumbersome purification. Furthermore, Nevertheless, CEPS is currently rarely applied in aca-
chemical peptide segment condensation can cause demia and industry since it is not broadly applicable
racemization, especially at the coupling site, and re- and suffers from severe hydrolytic side-reactions. An-
Adv. Synth. Catal. 0000, 000, 0 – 0
1
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
other available coupling method is native chemical li- actions without (hydrolytic) side reactions are highly
[5]
gation which, however, relies on the presence of cer- desired. Since such enzymes have not been discov-
tain amino acids in the peptide sequence, e.g., an N- ered, a stabilized and cation (calcium) independent
[
15]
terminal Cys, and is not scalable due to the instability variant of subtilisin BPN’ (Sbt149)
of peptide thioesters. our study as the starting point for protein engineering.
CEPS is usually carried out by using (hydrolytic) This variant has a deleted calcium binding domain
was chosen in
[
6]
proteases in the reverse direction. Since peptide (deletion of 9 amino acids, number 75 to 83) and an
bond synthesis is thermodynamically unfavourable in additional 18 stabilizing mutations, including one di-
the presence of water, the use of neat organic solvents sulfide bridge. We herein report that by introducing
with very low water activity can be beneficial for pep- the Ser212Cys and the Pro216Ala mutations into this
tide coupling. For instance, the industrial serine pro- robust enzyme scaffold, a new subtilisin variant is ob-
tease Alcalase has been applied for peptide synthesis tained, termed peptiligase, that catalyzes efficient
[
7]
under nearly anhydrous conditions. However, pro- peptide ligation in water.
teases exhibit very low activity in neat organic sol-
[8]
vent, and the peptide substrates are often poorly
soluble both when side chains are (partially) protect- Results and Discussion
ed. An alternative approach would be performing
coupling reactions in water. Here, a peptide that is C- The gene encoding peptiligase was constructed by
terminally activated as ester (acyl donor) is coupled site-directed mutagenesis and cloned into the Escheri-
to a peptide N-terminal amine (acyl acceptor), and chia coli/Bacillus subtilis shuttle vector pBS42. Pro-
the enzyme or reaction system is designed to kineti- duction of the enzyme was performed in B. subtilis
cally favour formation of the ligation product. There strain DB104 which lacks native extracellular neutral
[
16]
are some effective ligating enzymes known from and serine proteases.
Peptiligase is translated as
[
9]
nature, such as sortase and butelase, but they re- a precursor and needs to be processed to obtain the
quire specific recognition sequence motifs. Another mature protein. It was reported that a ꢂhelperꢀ pro-
approach is the use of substrate mimetics and a pro- tease, i.e., native subtilisin, is required for processing
[
10]
[12]
tease. However, synthesis of the required highly acti- of subtiligase, due to its reduced protease activity.
vated peptide esters is challenging and some amino In the present study, we observed that peptiligase can
acids may undergo side reactions, such as Cys, Met, perform the autoproteolytic removal of the pro-se-
[11]
Trp, His and Tyr.
quence without any additional subtilisin, which sug-
Protein engineering has been used to tailor more gests that the retained proteolytic activity of this
broadly applicable proteases for peptide coupling in enzyme is sufficient for its maturation. Secreted pepti-
[
12]
water. Wells and co-workers
constructed subtili- ligase was purified from the culture medium by am-
gase, a double mutant derived from the Bacillus amy- monium sulfate precipitation and ion-exchange chro-
loliquefaciens subtilisin BPN’. The catalytic serine res- matography. The yield from 1 liter of cell culture was
idue was mutated to a cysteine, and a proline near the approximately 20 mg (70% pure on SDS-PAGE gel).
active site was converted to an alanine to reduce The thermostability of the purified peptiligase was de-
steric crowding. The hydrolytic activity of this mutant termined using the thermofluor assay. The apparent
is significantly reduced compared to the wild-type, transition temperature (Tm,app) of 668C indicates that
since the introduced cysteine forms a thioester inter- the mutated enzyme well preserves the thermostabili-
mediate with the acyl donor, which is more prone to ty of its parent subtilisin.
the amine nucleophilic attack leading to the ligation
As anticipated, peptiligase is cation-independent. In
product. Unfortunately, the average ligation yield is a standard peptide ligation reaction consisting of the
[12a]
6
6%,
acceptor,
non-viable. Elliott
using a mutant of Streptomyces griseus protease B cations tested (Ca , Mg , Mn , Ni ; 10 mM con-
SGPB) for peptide coupling. Here, the active site centration) displayed a remarkably stimulatory or in-
even when using a ten-fold excess of the acyl activated carboxamidomethyl (Cam) ester Ac-Phe-
[12b]
making this technology economically Ile-Glu-Trp-Leu-OCam as acyl donor and H-Ala-Phe-
[
13]
reported another approach by NH as the amidated acyl acceptor none of the metal
2
2+
2+
2+
2+
(
catalytic serine is mutated to an alanine and the reac- hibitory effect on the activity of peptiligase. The influ-
tion proceeds through a proposed histidine-involved ence of the metal chelator EDTA (10 mM) on the en-
acyl-enzyme intermediate. The stability of both en- zymatic activity was also negligible. In addition, the
zymes was improved by introducing additional muta- thermostability of peptiligase was unaffected by
tions, yielding stabiligase and streptoligase, respective- metals or EDTA (Table 1). The pH-dependence was
[14]
ly. However, in both cases, an enzyme bound cation determined in MES and Tricine buffers with pH
is crucial for the stability, which restricts their applica- values ranging from 6.5–8.85. The optimal pH value
tion. Alternative broadly applicable, robust and (highest synthesis/hydrolysis ratio) was pH 8.0. When
cation-free enzymes that catalyze peptide coupling re- a higher pH was used, chemical hydrolysis of the pep-
Adv. Synth. Catal. 0000, 000, 0 – 0
2
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
Table 1. Effect of metal ions and chelating agent EDTA on tide ester was more profound. At lower pH, the
the
(Ac-Phe-Ile-Glu-Trp-Leu-OCam+H-Ala-Phe-NH2)
amine acyl acceptor is unreactive due to protonation.
To test whether peptiligase tolerates high concen-
trations of organic solvents, peptide ligation reactions
were performed in the presence of various amounts
of water-miscible cosolvents, including tetrahydrofur-
an (THF), N,N-dimethylformamide (DMF) and di-
methyl sulfoxide (DMSO). Reaction mixtures again
contained the Ac-Phe-Ile-Glu-Trp-Leu-OCam acyl
coupling activity and thermostability of peptiligase. Details
are given in the Experimental Section.
Activity [mU/mg]
Tm, app [8C]
control
26
20
26
30
22
26
66.0
65.5
65.0
64.5
62.0
66.5
2
+
Ca (10 mM)
2+
Mg (10 mM)
2+
Mn (10 mM)
2+
donor and H-Ala-Phe-NH as acceptor. Conversion of
2
Ni (10 mM)
substrate and product formation were estimated by
HPLC.
EDTA (10 mM)
As shown in Figure 1, the enzyme exhibited moder-
ate tolerance towards THF, and a reasonable residual
Figure 1. Effect of organic solvents and denaturing agent on peptiligase-catalyzed peptide coupling of Ac-Phe-Ile-Glu-Trp-
Leu-OCam and H-Ala-Phe-NH . Black bars: relative total activity; grey bars: relative synthetic activity; grey triangles: log
2
value of synthesis/hydrolysis ratio. Details are given in the Experimental Section.
Adv. Synth. Catal. 0000, 000, 0 – 0
3
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
total activity (of 30%) was observed in the presence the acyl acceptor substantially occupying the binding
of 20% (v/v) THF. However, THF strongly sup- pockets of the enzyme, we used substrates of varying
pressed the nucleophilicity of the acyl acceptor and lengths in coupling reactions (Table 2, Table 3 and
hydrolysis of the acyl donor became dominant. Only Table 4). First, single amino acids and dipeptides were
hydrolysis of the acyl donor Cam ester to free acid coupled via a kinetically controlled conversion. Cam-
was observed. In contrast, the S/H ratio was not sig- esters are optimal substrates for proteases used as li-
[
12]
nificantly affected by the addition of either DMF or gases as they mimic a glycine residue.
Peptide
DMSO. The enzyme also showed good stability in Cam-esters are easily produced via standard solid
these two cosolvent systems. Notably, the enzyme re- phase peptide synthesis using conventional resins in
[
18]
tained 36% of its activity in the presence of 50% (v/ high yield and good purity.
However, when such
v) DMSO in water, which is a very good mixture for amino acids and dipeptides were tested in coupling
dissolving hydrophobic peptides. Therefore, if re- reactions, the efficiency by peptiligase was poor
quired for substrate solubility, DMSO is the preferred (Table 2).
organic cosolvent to use in coupling reactions.
A possible issue with peptide ligation reactions, as acyl donor Cam-esters with the same ligation junc-
Subsequently, a series of longer (5–11 amino acids)
[
14a]
described by Chang et al.,
is that substrates could tions were tested. The results indicate that peptiligase
adopt a secondary structure reducing the accessibility catalyzes the coupling of longer acyl donors with high
to the binding pockets of the coupling enzyme. In efficiency, and that there is a clear correlation be-
such cases, denaturing agents like guanidinium hydro- tween the length of the acyl donor and the S/H ratio
chloride (GndCl) or urea can be added to weaken the of the reaction (Table 3). Virtually no hydrolytic side-
secondary structure and expose the ligation junction reactions were observed at all when an 11-mer Cam-
of the substrates. Under extreme denaturing condi- ester was used, i.e., the S/H ratio exceeded 100. This
tions, peptiligase retained significant activity; up to corresponds to a conversion of >99% to product and
4
M GndCl could be used. Moreover, the S/H ratio less than 1% of hydrolysis. Notably, this high yield
was only marginally influenced by the addition of was obtained using only 1.5 equivalents of acyl ac-
GndCl.
ceptor, making this an economically attractive liga-
Next, we investigated whether increasing the length
Next, the efficiency of peptide ligation with peptili- tion.
gase was investigated. The parent enzyme Sbt149
binds the substrate in an extended antiparallel b-sheet of the acyl acceptor has a similar beneficial effect
conformation, while substrate residues surrounding (Table 4). The efficiency of peptide ligation was ana-
the scissile bond (peptide groups P4–P3’) are in close lyzed using a series of acyl acceptors with the same
contact with their respective binding pockets (subsites amino-terminal residue and stepwise length extension
[17]
S4–S3’). To test whether ligation reactions catalyzed (1–4 amino acids). While a coupling reaction with ala-
by peptiligase also require both the acyl donor and nine amide as nucleophile appeared to be accompa-
[a]
Table 2. Peptiligase-catalyzed coupling reactions of amino acids and dipeptides .
Acyl donor
Acyl acceptor
Conversion [%]
Synthesis [%]
Hydrolysis [%]
S/H ratio
Cbz-F-OCam
Cbz-F-OCam
Cbz-AF-OCam
Cbz-AF-OCam
H-A-NH2
H-AF-NH2
H-A-NH2
H-AF-NH2
26
50
53
53
3
6
2
6
23
44
51
47
0.13
0.13
0.04
0.13
[
a]
Reactions were performed with 1 mM acyl donor and 3 mM acyl acceptor in 100 mL Tricine buffer (100 mM, pH 8.0). Re-
action time = 3 h.
Table 3. Peptiligase-catalyzed coupling reactions using acyl donors with varying lengths and H-AF-NH as the amine nucleo-
2
[
a]
phile .
Acyl donor
Conversion [%]
Synthesis [%]
Hydrolysis [%]
S/H ratio
Ac-DLSKQ-OCam
93
85
66
60
81
83
64
60
12
2
2
6
41.5
32
Ac-TSDLSKQ-OCam
Ac-TFTSDLSKQ-OCam
Ac-EGTFTSDLSKQ-OCam
<1
>100
[
a]
Reactions were performed with 10 mM acyl donor and 15 mM acyl acceptor in 100 mL Tricine buffer (100 mM, pH 8.0).
Reaction time=3 h.
Adv. Synth. Catal. 0000, 000, 0 – 0
4
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
Table 4. Peptiligase-catalyzed coupling reactions using amine nucleophiles with varying lengths and Ac-FIEWL-OCam as
[a]
the acyl donor .
Acyl acceptor
Conversion [%]
Synthesis [%]
Hydrolysis [%]
S/H ratio
H-A-NH2
35
59
25
23
<1
32
13
35
27
11
9
<0.01
1.2
1.2
H-AF-NH2
H-AFA-NH2
H-AFAY-NH2
14
1.5
[
a]
Reactions were performed with 1 mM acyl donor and 3 mM acyl acceptor in 100 mL Tricine buffer (100 mM, pH 8.0). Re-
action time=1 h.
nied by overwhelming hydrolysis, longer peptide gase; in most cases the S/H ratio of peptiligase was
amides served as better acyl acceptors.
two-fold higher.
These results suggest that occupation of both the
The improved S/H ratio together with the excellent
S1’ and the S2’ subsites as well as the S1–S4 subsites stability render peptiligase a highly promising enzyme
is a prerequisite for efficient ligation. Thus, the mini- for further development of a generally applicable
mal peptide length for an efficient coupling reaction CEPS technology. We are currently investigating the
is 4+2, with longer segments giving more efficient li- peptiligase-catalyzed synthesis of various pharmaceut-
gation, as desired. From the results of an 11+8 cou- ical products on a gram scale to determine purified
pling reaction, i.e., Ac-Lys-Lys-Lys-Lys-Lys-Lys-Asp- product yield and compare the results to those of con-
Phe-Ser-Lys-Leu-OCam-Leu-OH+H-Ala-Ala-Pro-
ventional synthetic methods.
Arg-Ala-Ala-Arg-Glu-OH, it became clear that we
Besides peptide segment condensation, subtiligase
have not reached the limitations of this technology has been used to cyclize linear peptides longer than
since the reaction reached 97% conversion to a 19- 13 amino acids in reasonable yield (average of
[
19]
mer product with only 3% hydrolysis of the ester 64%).
bond (details in the Supporting Information). Clearly, class of pharmaceuticals, mainly due to their high sta-
Macrocylic peptides are a very promising
[
20]
enzymatic ligation can be used for the synthesis of bility in blood serum. We embarked on the cycliza-
both medium-sized and long peptides. tion (Figure 2) of linear Microcin J25, the precursor
To determine whether the synthetic properties of of a 21 amino acid long lasso-peptide with antibiotic
peptiligase are better than those of subtiligase, S/H properties that is naturally produced by a strain of E.
[
21]
ratios were determined using various acyl acceptors coli isolated from human feces. Chemical head-to-
Table 5). Sub-optimal (diluted) reaction conditions tail cyclization of such a long peptide is very challeng-
(
were used to clearly quantify the amounts of synthetic ing and the reaction has to be performed with ex-
product and product of the hydrolytic side reaction. tremely diluted substrate to prevent polymerization,
[
22]
Under the same reaction conditions, the S/H ratio of making it difficult for scale-up. Microcin J25 is hy-
peptiligase is surprisingly higher than that of subtili- drophobic and cosolvents such as DMSO are needed
Table 5. P1’ and P2’ substrate scope of peptiligase and subtiligase using Ac-DFSKL-OCam as acyl donor and various acyl ac-
[a]
ceptors, where Xxx=the proteinogenic amino acid indicated. Best eight results shown .
Acyl acceptor
H-Xxx-Leu-Arg-NH2
Peptiligase
Synthesis [%]
Subtiligase
Synthesis [%]
[b]
[b]
S/H ratio
S/H ratio
Ser
58
48
47
21
1.38
0.92
0.89
0.27
40
25
20
6.8
0.67
0.33
0.25
0.07
Gly
Ala
Asp
[b]
[b]
H-Ala-Xxx-Arg-NH2
Synthesis [%]
S/H ratio
Synthesis [%]
S/H ratio
Ile
46
46
32
31
0.83
0.86
0.46
0.38
26
29
26
24
0.35
0.41
0.34
0.28
Val
Leu
Met
[
a]
Reactions were performed with 0.75 mM acyl donor and 1.5 mM acyl acceptor in 1 mL Tricine buffer (100 mM, pH 8.0)
and 5 mg of enzyme. Full conversion was reached after 30 min.
Product yields were derived from HPLC analysis of the reaction mixtures.
[
b]
Adv. Synth. Catal. 0000, 000, 0 – 0
5
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
Figure 2. Head-to-tail cyclization of linear Microcin J25 (21-mer) by peptiligase in 20% (v/v) DMSO (details in the Experi-
mental Section).
in high amounts to solubilize the peptide in aqueous Experimental Section
environment, which is well tolerated by peptiligase.
When Cam-Leu-activated linear Microcin J25 Construction of Peptiligase Expression Plasmid
À1
(
(
1 mgmL ) was cyclized with peptiligase using 20%
v/v) DMSO in phosphate buffer, a very good conver-
The Sbt149 encoding gene with its natural promoter se-
quence (a kind gift from Prof. P. N. Bryan, University of
Maryland, USA) was used as the template to generate the
Ser212Cys/Pro216Ala mutant by the megaprimer method.
sion to cyclic product was observed, i.e., 82% of syn-
thetic product and only 18% of hydrolysis of the
linear Microcin J25Cam-Leu ester to produce the free The final PCR product was purified, digested with EcoRI
acid. No dimerization or polymerization was detected. and BamHI, and ligated into the EcoRI/BamHI treated
Moreover, a similar conversion to cyclic peptide pBS42 vector (DSM8748, DSMZ, Braunschweig, Germany).
The ligation mixtures were transformed to competent E.
(
81%) was obtained in a reaction with concentrated
À1
coli ER 1821 cells (NEB, Ipswich, MA, USA) and trans-
formants were plated on LB plates containing 34 mgmL
chloramphenicol. The mutant gene was confirmed by DNA
sequencing.
substrate (100 mgmL in DMSO) dosed to peptili-
gase in phosphate buffer, thus obtaining a concentrat-
ed product solution (50 mgmL ). Clearly, besides
peptide segment condensation, peptiligase can be
used to synthesize head-to-tail macrocyclic peptides
in an efficient manner.
À1
À1
Expression and Purification of Peptiligase
Plasmid pBS42-peptiligase was transformed into B. subtilis
DB104, which is a double mutant (his nprR2 nprE18
Conclusions
[23]
aprD3) strain deficient in extracellular neutral and serine
proteases and was a kind gift from Prof. O. Kuipers (Univer-
sity of Groningen, the Netherlands). Transformants grown
We have successfully constructed a stable peptide
ligase that is able to ligate peptides in water with very
high S/H ratios. When longer peptides are coupled,
À1
on an LB plate containing 10 mgmL chloramphenicol at
the hydrolytic side-reaction is almost completely sup- 378C for 16 h were picked and inoculated into 10 mL of LB
À1
pressed. To the best of our knowledge, this is the first broth containing 10 mgmL chloramphenicol. After 16 h of
reported broadly applicable and traceless peptide li- incubation at 378C, 1% (v/v) of the cultures were inoculated
À1
À1
À1
to 1 L rich broth (20 gL yeast extract, 5 gL tryptone,
gating enzyme that is independent of cations and is
very robust. The enzyme can be expressed in B. subti-
lis and can be directly isolated from the culture
medium since it is secreted and no additional protease
is required for maturation. Furthermore, peptiligase is
compatible with a variety of additives and cosolvents,
has a relaxed specificity in coupling reactions, and
6
5
.4 g NaCl, 30 mM phosphates pH 7.6, 10 gL glucose,
À1
À1
À1
0 mgL
Trp, 50 mgL
Lys, 50 mgL
Met, 0.06 mM
MnCl ). Cultures were grown at 378C with shaking for 48 h.
2
Next, the medium was harvested by centrifugation at 6,000 g
for 20 min at 48C. Subsequently, 5 g of CaCl were added to
2
the medium to precipitate phosphate salts and the pH was
adjusted to 7.5. The precipitate was removed by centrifuga-
does not hydrolyze amide bonds in the main chain of tion at 6,000 g for 20 min, 48C. Ammonium sulfate was
precursor or product peptides. These features endow added to the supernatant to a final concentration of 45 wt%.
the enzyme with the catalytic properties required for The mixture was stirred for 1 h at 48C to precipitate peptili-
gase, which was harvested by centrifugation at 8,000 g for
the synthesis of a variety of medium-sized and long
3
0 min at 48C. The pellet was washed 4 times with 100 mL
(
pharmaceutical) peptides. Moreover, we have shown
8
0% acetone, and resuspended in 15 mL water. Resuspend-
that peptiligase can be used for the synthesis of a pep-
tide macrocycle, which represents an important up-
coming class of pharmaceutical peptides for future
medicines. Based on the robust framework of peptili-
gase, further protein engineering for even higher syn-
ed material was centrifuged at 15,000 g for 10 min at 48C, to
remove insoluble material. The supernatant was desalted
using a HiPrep 26/10 desalting column in buffer (20 mM Tri-
cine, pH 7.5). The desalted proteins were loaded on a 5 mL
HiTrap Q HP column. The flow-through containing peptili-
thetic performance or broader substrate acceptance is gase was collected and concentrated. The purity of the pro-
ongoing.
tein was analyzed by SDS-PAGE and estimated to be 70%.
Adv. Synth. Catal. 0000, 000, 0 – 0
6
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
The purified enzyme was flash-frozen in liquid nitrogen and
stored at À808C until further use.
7.5–8.85). The chemical hydrolysis of the acyl donor was as-
sayed in the absence of the enzyme.
Effect of Organic Solvent and Additives on
Peptiligase Activity
Expression and Purification of Subtiligase
[12]
A gene encoding subtiligase
was provided by DNA2.0
The standard reaction mixtures consisted of 1 mM Ac-Phe-
Ile-Glu-Trp-Leu-OCam, 3 mM Ala-Phe-amide, 15 mM
enzyme and different concentrations of organic solvents,
metal ions, chelating agent or denaturing agent in 100 mM
Tricine buffer (pH 8.0). The assays were performed at 258C.
The uncatalyzed chemical hydrolysis of the acyl donor was
assayed in the absence of the enzyme.
(
https://www.dna20.com/) and recloned into EcoRI-BamHI
digested E. coli–B. subtilis shuttle vector pBS42. The plas-
mid pBS42-S5 was propagated in E. coli strain MM294
(
validated by sequencing. The validated plasmid was used for
transformation of B. subtilis DB104. The production and pu-
rification protocol was similar to that for peptiligase, but
DSM5208, DSMZ, Braunschweig, Germany), isolated and
using buffers supplemented with 1 mM CaCl . The yield of
purified subtiligase was 2–5 mgL culture.
2
À1
Synthesis of a 19 Amino Acid Peptide
The coupling reaction was performed using 4 mM acyl
donor (Ac-Lys-Lys-Lys-Lys-Lys-Lys-Asp-Phe-Ser-Lys-Leu-
OCam-Leu-OH) and 6 mM acyl acceptor (H-Ala-Ala-Pro-
Arg-Ala-Ala-Arg-Glu-OH) in phosphate buffer (1M,
pH 8.0). To this reaction mixture, 5 mg of enzyme was added
and the mixture (1 mL) was analyzed by LC-MS after
Thermostability Assays
[24]
The fluorescence based thermal stability assay was used
to determine apparent melting temperatures of the proteins.
For this, 20 mL of protein solution in buffer (20 mM Tricine,
pH 7.5) were mixed with 5 mL of 100 x Sypro Orange (Mo-
lecular Probes, Life Technologies, San Diego, CA) dye in
a thin wall 96-well PCR plate. The plate was sealed with
Optical-Quality Sealing Tape and heated in a CFX 96 Real
Time PCR System (BioRad, Hercules, CA, USA) from 20
3
0 min. The Cam-ester starting material was fully consumed
and the ratio between the product and hydrolysis peak was
7/3. See the Supporting Information for further details.
9
À1
Cyclization of Linear Microcin J25 Cam-Leu-Ester
to 998C at a heating rate of 1.758Cmin . Fluorescence
changes were monitored with a charge-coupled device
The cyclization reaction was performed with 1 mg linear
peptide Cam-Leu ester (H-Gly-Thr-Pro-Ile-Ser-Phe-Tyr-Gly-
Gly-Gly-Ala-Gly-His-Val-Pro-Glu-Tyr-Phe-Val-Gly-Ile-
OCam-Leu-OH) in 1 mL phosphate buffer (100 mM,
pH 8.0) supplemented with DMSO (20 vol%). To this reac-
tion mixture, 10 mg of peptiligase were added and the solu-
tion was analyzed by LC-MS after 30 min. Conversion to
product was measured by integrating the starting material,
product and hydrolyzed Cam-ester peaks. The amount of
product after 30 min was 82% and the amount of hydrolysis
18%. The Cam-ester starting material was completely con-
sumed. See the Supporting Information for further details.
In a second reaction, concentrated substrate solution in
(
CCD) camera. The wavelengths for excitation and emission
were 490 and 575 nm, respectively.
Peptide Ligation Reactions
The ligation reactions were typically performed at 258C in
100 mM Tricine buffer (pH 8.0), containing 15 mM peptili-
gase and different acyl donors and acyl acceptors. The effi-
ciency of the ligation reactions was analyzed by an LC/MS
ion-trapping system (Thermo Scientific). The reaction mix-
tures were loaded on an Alltech C18 3u column (3 mm,
100 mmꢃ4.6 mm). The solvent system consisted of 0.01%
À1
formic acid (eluent A) and 0.08% formic acid in 30% aceto-
nitrile (eluent B). A typical elution gradient was as follows:
start with 100:0 A:B for 10 min, in 45 min from 100:0 to
DMSO (100 mgmL ) was prepared and dosed in time
(20 mL every 15 min) to the enzyme solution (10 mg peptili-
gase) in 200 mL phosphate buffer (100 mM, pH 8.0). The pH
was continuously kept at 8.0 using 5 N aqueous NaOH. The
reaction mixture was analyzed by LC-MS after 150 min as
described above. The starting material was fully consumed
and the amount of product was 81% and hydrolysis 19%.
40:60, and from 40:60 to 100:0 from 45 min to 60 min. The
analyses were carried out at 258C, with concomitant UV de-
tection at 220 nm or 280 nm. For reactions using tryptophan-
containing acyl donors, the UV detection was set at 280 nm,
and the concentrations of the compounds were calculated
based on their peak area. For reactions using tryptophan-
free acyl donors, the UV detection was set at 220 nm and
the conversion was calculated based on the consumption of
the acyl donor substrate. Chemical hydrolysis of the acyl
donor was calculated from control experiments.
Acknowledgements
This project is part of Integration of Biosynthesis and Organ-
ic Synthesis program (IBOS-2; program number: 053.63.014)
funded by The Netherlands Organisation for Scientific Re-
search (NWO) and Advanced Chemical Technologies for
Sustainability (ACTS). The authors thank Dr. J. M. van der
Laan from DSM Food Specialties (Delft, The Netherlands)
for helpful discussions and A. C. van de Meulenreek, B.Sc.
M.B., for the synthesis of linear Microcin J25 Cam-Leu-OH
ester. Authors declaration: AT, TN, and PJMLQ are em-
pH-Ligation Profile
The reaction mixture containing 1 mM Ac-Phe-Ile-Glu-Trp-
Leu-OCam, 3 mM H-Ala-Phe-amide, and 15 mM peptiligase
were incubated at 258C for 1 h. To determine the effect of
pH on the ligation efficiency, experiments were carried out
over a pH range of 6.0–8.85. The buffers used were MES
(
100 mM, pH range 6.0–7.0) and Tricine (100 mM, pH range
Adv. Synth. Catal. 0000, 000, 0 – 0
7
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
asc.wiley-vch.de
ployed by Enzypep B. V. Some of Enzypepꢀs products are
based on the use of protease technology.
[10] a) N. Wehofsky, A. Pech, S. Liebscher, S. Schmidt, H.
Komeda, Y. Asano, F. Bordusa, Angew. Chem. 2008,
1
20, 5536–5540; Angew. Chem. Int. Ed. 2008, 47, 5456–
5
460; b) N. Wehofsky, S. Thust, J. Burmeister, S. Kluss-
mann, F. Bordusa, Angew. Chem. 2003, 115, 701–704;
Angew. Chem. Int. Ed. 2003, 42, 677–679.
11] Y. Yang, Side Reactions in Peptide Synthesis, Academic
References
[
[
1] a) A. M. Thayer, Chem. Eng. News 2011, 89, 13–20;
b) A. M. Thayer, Chem. Eng. News 2011, 89, 21–25.
2] a) H. Korhonen, A. Pihlanto, Int. Dairy J. 2006, 16,
Press, Cambridge, 2016.
[12] a) A. Braisted, J. Judice, J. A. Wells, Methods Enzymol.
1997, 289, 298–313; b) L. Abrahmsen, J. Tom, J. Burni-
er, K. A. Butcher, A. Kossiakoff, J. A. Wells, Biochem-
istry 1991, 30, 4151–4150.
[
9
45–960; b) H. Korhonen, A. Pihlanto, Curr. Pharm.
Des. 2007, 13, 829–843.
[
[
[
[
3] D. Hans, P. R. Young, D. P. Fairlie, Med. Chem. 2006, 2,
[13] R. J. Elliott, A. J. Bennet, C. A. Braun, A. M. Ma-
cLeod, T. J. Borgford, Chem. Biol. 2000, 7, 163–171.
[14] a) T. K. Chang, D. Y. Jackson, J. P. Burnier, J. A. Wells,
Proc. Natl. Acad. Sci. USA 1994, 26, 12544–12548;
b) K. Joe, T. J. Borgford, A. J. Bennet, Biochemistry
2004, 43, 7672–7677.
6
27–646.
4] F. Guzman, S. Barberis, A. Illanes, EJB Electron. J.
Biotechnol. 2007, 10, 279–314.
5] E. Dawson, T. W. Muir, I. Clark-Lewis, S. B. Kent, Sci-
ence 1994, 266, 776–779.
6] a) J. Bongers, E. P. Heimer, Peptides 1994, 15, 183–193;
b) F. Bordusa, Chem. Rev. 2002, 102, 4817–4867; c) S.
Nakano, S. Okazaki, E. Ishitsubo, N. Kawahara, H.
Komeda, H. Tokiwa, Y. Asano, Sci.Rep. 2015, 5, 13836;
d) K. Yokozeki, S. Hara, J. Biotechnol. 2005, 115, 211–
[15] B. Ruan, V. London, K. E. Fisher, D. T. Gallagher, P. N.
Bryan, Biochemistry 2008, 47, 649–656.
[16] A. Longo, J. R. Snay, L. Zimmer, M. M. Ataai, Biotech-
nol. Lett. 1988, 10, 649–656.
[17] S. L. Strausberg, B. Ruan, K. E. Fisher, P. A. Alexand-
er, P. N. Bryan, Biochemistry 2005, 44, 3272–3279.
[18] R. J. A. C. de Beer, T. Nuijens, L. Wiermans, P. J. L. M.
Quaedflieg, F. P. J. T. Rutjes, Org. Biomol. Chem. 2012,
10, 6767–6775.
2
20.
[
7] a) T. Nuijens, A. C. H. M. Schepers, C. Cusan, J. A. W.
Kruijtzer, D. T. S. Rijkers, R. M. J. Liskamp, P. J. L. M.
Quaedflieg, Adv. Synth. Catal. 2013, 355, 287–293; b) T.
Nuijens, C. Cusan, T. J. G. M. van Dooren, H. M.
Moody, R. Merkx, J. A. W. Kruijtzer, D. T. S. Rijkers,
R. M. J. Liskamp, P. J. L. M. Quaedflieg, Adv. Synth.
Catal. 2010, 352, 14–15; c) T. Nuijens, C. Cusan,
A. C. H. M. Schepers, J. A. W. Kruijtzer, D. T. S. Rijk-
ers, R. M. J. Liskamp, P. J. L. M. Quaedflieg, J. Mol.
Catal. B 2011, 71, 79–84; d) T. Nuijens, E. Piva, J. A. W.
Kruijtzer, D. T. S. Rijkers, R. M. J. Liskamp, P. J. L. M.
Quaedflieg, Adv. Synth. Catal. 2011, 353, 1039–1044;
e) P. Vossenberg, H. H. Beeftink, T. Nuijens, M. A. C.
Stuart, J. Tramper, J. Mol. Catal. B 2012, 75, 43–49.
8] A. M. Klibanov, Trends Biotechnol. 1997, 15, 97–101.
9] a) M. Ritzefeld, Chemistry 2014, 20, 8516–8529;
b) G. K. T. Nguyen, S. Wang, Y. Qiu, X. Hemu, Y. Lian,
P. J. Tam, Nat. Chem. Biol. 2014, 10, 732–738.
[19] D. Y. Jackson, J. P. Bumier, J. A. Wells, J. Am. Chem.
Soc. 1995, 117, 819–820.
[20] S. M. Howell, S. V. Fiacco, T. T. Takahashi, F. Jalali-
Yazdi, S. W. Millward, B. Hu, P. Wang, R. W. Roberts,
Sci. Rep. 2014, 4, 1–5.
[21] K. A. Wilson, M. Kalkum, J. Ottesen, J. Yuzenkova,
B. T. Chait, R. Landick, T. Muir, K. Severinov, A.
Darst, J. Am. Chem. Soc. 2003, 125, 12475–12483.
[22] L. Z. Yan, P. E. Dawson, J. Am. Chem. Soc. 2001, 123,
526–533.
[23] F. Kawamura, R. H. Doi, J. Bacteriol. 1984, 160, 442–
444.
[24] U. B. Ericsson, H. M. Hallberg, G. T. DeTitta, N.
Dekker, P. Nordlund, Anal. Biochem. 2006, 357, 289–
298.
[
[
Adv. Synth. Catal. 0000, 000, 0 – 0
8
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!

FULL PAPERS
Peptiligase, an Enzyme for Efficient Chemoenzymatic
Peptide Synthesis and Cyclization in Water
9
Adv. Synth. Catal. 2016, 358, 1 – 9
Ana Toplak, Timo Nuijens, Peter J. L. M. Quaedflieg,
Bian Wu,* Dick B. Janssen*
Adv. Synth. Catal. 0000, 000, 0 – 0
9
ꢁ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÞÞ
These are not the final page numbers!