Generation of active protease depending on peptide-protein interactions using interaction-dependent native chemical ligation and protein trans-splicing

Article abstract of DOI:10.1246/bcsj.20190159

An artificial signal transduction system has been constructed by employing engineered human immunodeficiency type-1 (HIV-1) protease and Nostoc punctiforme PCC73102 (Npu) DnaE intein. While the truncation of four amino acid residues at the N-terminus of H

Full text of DOI:10.1246/bcsj.20190159

Advance Publication Cover Page
Generation of Active Protease Depending on Peptide-Protein Interactions Using
Interaction-Dependent Native Chemical Ligation and Protein Trans-Splicing
Tsuyoshi Takahashi
Advance Publication on the web July 9, 2019
doi:10.1246/bcsj.20190159
© 2019 The Chemical Society of Japan
Advance Publication is a service for online publication of manuscripts prior to releasing fully edited, printed versions. Entire
manuscripts and a portion of the graphical abstract can be released on the web as soon as the submission is accepted. Note that the
Chemical Society of Japan bears no responsibility for issues resulting from the use of information taken from unedited, Advance
Publication manuscripts.

Generation of Active Protease Depending on Peptide-Protein Interactions Using
Interaction-Dependent Native Chemical Ligation and Protein Trans-Splicing
Tsuyoshi Takahashi*1
1Chemistry and Chemical Biology, Graduate School of Science and Technology, Gunma University, 1-5-1, Tenjincho,
Kiryu, Gunma 376-8515
E-mail: <ttakahas@gunma-u.ac.jp>
Tsuyoshi Takahashi
Tsuyoshi Takahashi received Dr. Eng. from Tokyo Institute of Technology in 2001. In 2001-2002, he was a research fellowship of the
Japan Society for the Promotion of Science (JSPS) for Young Scientist. In 2002-2010, he was an Assistant Professor at Tokyo Institute of
Technology. At present he has been an Associate Professor at Gunma University from 2011. His research interests focus on peptide and
protein chemistry.
Abstract
An artificial signal transduction system has been
constructed by employing engineered human
To generate active enzymes and reporter proteins that
depend on input signals, such as binding events, protein-fragment
complementation assay has been widely employed.5-9 Sakamoto
and Kudo noted active green fluorescent protein (GFP) reporter
responses to host-guest interactions by employing split-GFP.7
Moreover, combining protein trans-splicing (PTS), that is
self-catalyzed by native and engineered split inteins,10-15 expands
its availability owing to the covalent bond formation between two
fragments (N-extein and C-extein) of the protein. Detection
immunodeficiency type-1 (HIV-1) protease and Nostoc
punctiforme PCC73102 (Npu) DnaE intein. While the
truncation of four amino acid residues at N-terminus of HIV-1
protease diminished its activity, the attachment of the PQIT
sequence into the truncated protease by protein trans-splicing
(PTS) reconstituted the enzymatic activity. By combining
interaction-dependent native chemical ligation (IDNCL) with
the PTS reaction, the peptide-protein interaction was clearly
detected by measuring HIV-1 protease activity. Src homology
domain 2 (SH2) of c-Src (SrcSH2) and phosphopeptides were
used as model binding pairs. HIV-1 protease activities were
dose-dependently increased after the IDNCL-PTS reaction
when the peptides containing pYEEI (pY = phosohotyrosine)
and pYEE sequences were used as the input peptides. HIV-1
protease activity generated by IDNCL-PTS might activate
several enzymes, and therefore, the artificial signal transduction
system might be available in synthetic biology.
methods
for
protein-protein
interactions
and
small
molecule-protein interactions have been developed using split
inteins with several enzymes and reporter proteins.16-19 Sonntag and
Mootz reported the generation of active tobacco etch virus (TEV)
protease by conditional protein splicing that depends on the
interaction of rapamycin with FK506 binding protein (FKBP) and
FKBP-rapamycin binding.19 In this system, intracellular proteins
can be manipulated through proteolysis by the active TEV protease.
Sonntag and Mootz used split VMA intein from Saccharomyces
cerevisiae in which two fragments (N-intein and C-intein) were
comprised of 184 and 64 amino acid residues, respectively. TEV
protease was also split into two pieces with more than 37 amino
acid residues. Therefore, it has been difficult to apply this method
to the development of a signal transduction system that depends on
the binding event between a synthetic ligand and a target protein,
because the synthetic ligand must be chemically conjugated to the
fusion protein comprising the TEV protease fragment and VMA
N-intein or C-intein.
Keywords:
peptide-protein
interaction,
protein
trans-splicing, HIV-1 protease, native chemical ligation,
signal transduction
1. Introduction
On the other hand, I have constructed a detection method for
ligand-protein interactions using interaction-dependent native
chemical ligation and PTS (IDNCL-PTS).20 In this method, IDNCL
reaction occurs between the short peptide (residues 24–34 of
β-galactosidase; βGal) bearing a ligand, such as maltose, via a
thioester linkage and N-intein of Ssp DnaB (residues 1–11) bearing
a target protein, such as maltose binding protein, when the ligand
binds to the target protein. Using this IDNCL product, active bGal
is generated by PTS with a fusion protein of C-intein of Ssp DnaB
M86 variant21 and C-terminal fragment of bGal (residues 35–1024).
Thus, the binding events were successfully transduced into the
βGal activity.
In nature, several signal transduction systems, such as kinase
cascades and ubiquitin signaling cascades, are used to maintain
higher-order functions in bio-organisms, especially in
eukaryotes.1,2 In these signal transductions, specific interactions
between receptor proteins and extracellular ligands usually work as
input signals. For instance, ligand binding to a death receptor that
induces apoptosis of the cell can activate an initiator caspase; it
then cleaves effector caspases that mediate several proteolytic
events of apoptosis. 3,4 Thus, construction of a mimetic system in
which binding events between a synthetic ligand and a target
protein is transduced as the generation of an active protease is
valuable.
In the present study, a signal transduction system that

activates a protease enzyme depending on the binding events has
been constructed. To do this, human immunodeficiency virus type
1 (HIV-1) protease was employed22 because it is a small protein
made up of two identical 99-amino acid polypeptide monomers and
it has catalytic activity when it forms a dimer.23,24 Split HIV-1
protease was designed and conjugated with an engineered split
intein, and the binding events between a ligand and a target protein
were transduced into active HIV-1 protease generated by
IDNCL-PTS.
protected peptide was obtained by a treatment with mixed
solvent of hexafluoro-2-propanol (HFIP) : dichloromethane =
1
:
4
and reacted with thiophenol (10 eq.) using
1-hydroxy-7-azabenzotriazole (3 eq.) and
diisopropylcarbodiimide (3 eq.). The protecting groups were
removed with TFA to give peptide thioester
(EELIKENMHMK-S-phenyl). C-terminal peptide
a
A
corresponding to the sequence of DnaE (1–35) was synthesized
on Fmoc-NH-SAL-PEG resin, and reacted by NCL with the
N-terminal peptide thioester to give Rpep-intN. The crude
peptides were purified by reversed-phase HPLC using a linear
gradient of acetonitrile/0.1% TFA. The peptides were identified
by their molecular ion peaks (M+H)+ by matrix-assisted laser
2. Experimental
Preparation of IntC-prC: The gene encoding HIV-1 protease
(5–99) was constructed by ligation between two
double-stranded DNA fragments, which were developed by
overlapping polymerase chain reaction (PCR) using synthetic
oligonucleotides. The gene encoding DnaE intein (36–137) was
amplified by PCR using pSKDuet16 plasmid (Addgene) as a
template.25 Two DNAs encoding HIV-1 protease and DnaE
intein were treated with MfeI/XhoI and KpnI/MfeI,
desorption/ionization–time-of-flight
mass
spectrometry
(MALDI–TOFMS; Shimadzu, AXIMA Performance): m/z
found (calcd): PrN-intN, 4428.32 (4428.31); Rpep-intN,
5371.50 (5371.73). The substrate peptide (Fsbst),
Abz-Thr-Ile-Nle-Phe(pNO2)-Gln-Arg-NH2,
was
also
synthesized on Rink Amide MBHA resin and identified by
RP-HPLC compared with that of the authentic sample,[27] and
MALDI–TOFMS; Fsbst, m/z found (calcd): 940.64 (940.50)
[(M+H)+].
respectively. The product DNA was ligated into
a
pET28a-based vector (Novagen) having a gene encoding YFP
as a fluorescent tag to give a plasmid, named pET28-intc-prc.
For the production of IntC-prC, BL21(DE3)pG·KJE8 was
transformed by pET22-intc-prc. The cells were grown in a
Luria-Bertani medium containing ampicillin as an antibiotic at
37°C to an optical density at 600 nm around 0.6. Expression of
Syntheses of PrN-pYEEI and PrN-pYEE: To prepare the
thioester
linkage,
N-(2,2-dimethoxyethyl)-2-[4-(mercaptomethyl)phenyl]acetami
de (compound 1) was synthesized according to Scheme S1.
Potassium thioacetate (2.05 g, 18.0 mmol) was added to a
solution of 4-(bromomethyl)phenylacetic acid (1.97 g, 8.60
mmol) in dimethylformamide (DMF; 25 mL). The resulting
solution was stirred at room temperature for one hour. DMF
was removed by evaporation, and the mixture was extracted
with ethyl acetate and 10% citric acid in water (x 1) and water
(x 2). The organic layer was dried over Na2SO4, concentrated
in vacuo, and recrystallized with ethyl acetate and hexane to
give compound 2. 1H NMR (400 MHz, CDCl3): d = 7.28 – 7.18
(m, 4 H), 4.09 (s, 2 H), 3.61 (s, 2 H), 2.33 (s, 3 H).
Dicyclohexylcarbodiimide (1.21 g, 5.89 mmol) was added to a
solution of compound 2 (1.20 g, 5.35 mmol) in tetrahydrofuran
(25 mL) on ice. After five minutes, aminoacetaldehyde
dimethyl acetal (0.58 mL, 5.35 mmol) was added and stirred
for one hour at 4°C, and for five hours at room temperature.
After filtration, the solvent was removed and the mixture was
extracted with ethyl acetate and 10% citric acid in water (x 2),
4% NaHCO3 (x 2), and brine (x 1). The organic layer was dried
over Na2SO4 and concentrated in vacuo. Silica gel
chromatography was performed with hexane and ethyl acetate
(2 : 3) to give compound 3 (0.97 g, 59%). Compound 3 was
treated with K2CO3 (1.3 eq.) in MeOH (30 mL) and water (10
mL) for one-and-a-half hours at room temperature. After
evaporation, ethyl acetate was added and extracted with 10%
citric acid in water (x 1), water (x 3), and brine (x 1). The
organic layer was dried over Na2SO4 and concentrated in vacuo.
Hexane was added, and the precipitate was collected by
the
protein
was
induced
by
0.4
mM
isopropyl-β-D-thiogalactopyranoside for 16 hours at 18°C. The
cells were harvested, and the expressed protein was extracted
by sonication. The protein was purified by Ni-NTA resin
(QIAGEN), and further purified by size-exclusion
chromatography using HiLoad 16/600 Superdex 75 column
(GE Healthcare) with a splicing buffer [50 mM Tris·HCl (pH
7.0), 300 mM NaCl, 1 mM EDTA, 2 mM DTT, 5% (v/v)
glycerol].
Preparation of IntN-SrcSH2: The gene encoding SrcSH2 was
amplified from a human placenta cDNA library (TaKaRa BIO)
by PCR. The product DNA was restricted with NdeI and XhoI
and ligated into a vector encoding Ssp DnaB intein (1–11) and
(Gly-Gly-Ser)12 linker. The region of the gene encoding DnaB
was replaced by double-stranded synthetic DNA encoding Npu
DnaE (1-35) to give a plasmid, named pET22-intn-sh2. The
production of IntN-SrcSH2 was performed using
BL21(DE3)pG·KJE8 as a host strain. After induction of protein
expression and harvesting cells, the protein was extracted from
the periplasmic space by osmotic shock using the TES buffer
[200 mM Tris·HCl (pH 8.0), 0.5 mM EDTA, 0.5 M sucrose].
The protein was purified by Ni-NTA resin, and the buffer was
exchanged using a desalting column (PD10, GE Healthcare) to
an assay buffer [100 mM Tris·HCl (pH 7.4), 100 mM NaCl].
Syntheses of PrN-intN and Rpep-intN: Peptide synthesis was
performed using the Fmoc (9-fluorenylmethyloxycarbonyl)
solid-phase
1
method
using
filtration to give compound 1 (587 mg, 70%). H NMR (400
O-(7-azabenzotriazole-1-yl)-1,1,3,3-tetramethyluronium
MHz, CDCl3): d = 8.13 (t, J = 5.3 Hz, 1H), 7.24-7.11 (m, 4H),
4.29 (t, J = 5.5 Hz, 1H), 3.65 (s, 2H), 3.20 (s, 6H), 3.11 (t, J =
5.5 Hz, 2H), 2.78 (s, 1H).
hexafluorophosphate as a coupling reagent. PrN-intN peptide
was elongated on an Fmoc-NH-SAL-PEG resin (Watanabe
Chemical Industry; 12.5 µmol). After elongation, peptidyl resin
was treated with trifluoroacetic acid (TFA) with 2.5% (v/v)
ethanedithiol, 2.5% (v/v) water, and 1% (v/v) triisopropyl
silane as scavengers. To synthesize Rpep-intN, two peptide
fragments were synthesized and conjugated by NCL. The
N-terminal peptide (EELIKENMHMK) was elongated on a
2-chlorotrityl chloride resin (Watanabe Chemical Industry; 36
µmol after incorporation of 1st Lys residue). A side-chain
A peptide thioester (PQIT-S-phenyl; 30 mg, 54.6 µmol)
and compound 1 (36 mg, 134 µmol) were dissolved in
trifluoroethanol (0.5 mL) and then 200 mM Tris·HCl buffer
(pH 8.0; 0.5 mL) was added and stirred for one-and-a-half
hours at room temperature (Scheme S2). Water (50 mL) was
added and evaporated. The crude mixture was purified by
RP-HPLC to afford PrN-CHO (9.1 mg, 21%). The peptide was
identified by MALDI–TOFMS; m/z found (calcd): 663.19

(663.32)
[(M+H)+].
Aoa-βA-pYEEI-NH2
and
is the difference between the value of the fluorescence increase
per hour in test solutions and in the absence of the peptide; and
ΔFmax is the difference between the value of fluorescence
increase per hour when IntN-SrcSH2 is completely complexed
with PrN-pYEEI or PrN-pYEE and in the absence of the
peptide.
Aoa-βA-pYEE-NH2 (Aoa = aminooxyacetyl) were synthesized
on Rink Amide MBHA resin (30 µmol). Ten µL of PrN-CHO
(10 mM in DMSO) and 10 µL of Aoa-βA-pYEEI-NH2 or
Aoa-βA-pYEE-NH2 (10 mM in DMSO) were mixed with 20
µL of 20 mM acetic acid in water. The reaction completion was
checked by RP-HPLC (Figure S1). MALDI–TOFMS;
PrN-pYEEI, m/z found (calcd): 1420.57 (1420.58) [(M+H)+];
PrN-pYEE, m/z found (calcd): 1307.72 (1307.50) [(M+H)+].
3. Results and Discussion
Design of split HIV-1 protease
In HIV-1 protease, N-terminal and C-terminal sequences,
residues 1–4 and 96–99, are crucial for dimer formation and
enzymatic activity because these residues of the two monomers
Splicing: Splicing reaction between Rpep-intN and IntC-prC
was performed in a splicing buffer at 25°C. [Rpep-intN] = 25
µM, [IntC-prC] = 5 µM. The assay was triplicated. The
reaction solution was analyzed by SDS-PAGE (15%
acrylamide), and image analysis was performed with the Image
J analysis software.
associate to form
a
four-stranded antiparallel b-sheet
structure.24 Therefore, N- or C-terminal deletion variants are
unable to form active dimeric structures. On this basis, I have
designed an engineered protein in which residues 1–4 of HIV-1
protease were replaced by a C-terminal fragment of Nostoc
punctiforme PCC73102 (Npu) DnaE (36–137)26 (see Figure 1).
Npu DnaE intein is a naturally occurring split intein11 that is
separated between residues 1–102 and 103–137. In my signal
Protease activity measurement: PrN-intN and IntC-prC were
reacted in a splicing buffer at 25°C. [PrN-intN] = 25 µM,
[IntC-prC] = 5 µM, [Fsbst] = 100 µM. Fluorescence spectra
were recorded on HITACHI F-2500 spectrofluorometer (λex =
335 nm). Fluorescence intensities at 440 nm were plotted. The
assay was triplicated.
transduction system,
a shorter peptide sequence of the
N-terminal fragment of the intein (N-intein) is better because it
needs to be placed at the N-terminus of the protein of interest.
Iwai and colleagues reported that an engineered Npu DnaE
intein whose sequence is divided between 1–35 and 36–137
was active.23 On this basis, I designed two engineered
molecules: IntC-prC protein having 36–137 residues of Npu
DnaE intein and 5–99 residues of HIV-1 protease, and
PrN-intN peptide having 1–4 residues of HIV-1 protease and
1–35 residues of Npu DnaE intein. The leucine residue at
position 5 in the HIV-1 protease was substituted by cysteine
because the cysteine residue is necessary at the +1 position of
the extein sequence in the Npu DnaE intein. A yellow
fluorescent protein (YFP) tag was also incorporated at the
N-terminus of IntC-prC.20 The IntC-prC protein was expressed
by an Escherichia coli (E. coli) strain, BL21(DE3)pG·KJE8,
and purified by size exclusion chromatography. The peptide
PrN-intN was synthesized using a solid-phase method and
purified by reversed-phase high performance liquid
IDNCL-PTS: In time-course experiments, PrN-pYEEI and
IntN-SrcSH2 were incubated in an assay buffer containing
tris(2-carboxyethyl)phosphine (TCEP) at 37°C for zero, two,
four, six, and eight hours. [PrN-pYEEI]
= 10 µM,
[IntN-SrcSH2] = 4 µM, [TCEP] = 2.5 mM. After incubation,
20 µL of each sample solution was mixed with 1 µL of Fsbst
(final conc. = 100 µM) and 21 µL of IntC-prC (final conc. =
2.6 µM). These were incubated at 25°C for eight hours. Then, 2
µL of 3 M NaOAc (pH 5.2) was added to the sample solution
to acidify, and the apparent dissociation constants of
PrN-pYEEI and PrN-pYEE with IntN-SrcSH2 were calculated
using equation (1) and a KaleidaGraph (Synergy Software), in
which a 1:1 binding stoichiometry was assumed.
ΔF = ΔFmax ([P]0 + [S]0 + Kd – (([P]0 + [S]0 + Kd)2
–
4([P]0[S]0)1/2)/(2[S]0)
chromatography (RP-HPLC). A peptide, Rpep-intN, for
(1)
checking the splicing activity was also designed and
synthesized (Figure 1).
where [P]0 and [S]0 represent the initial concentrations of
PrN-pYEEI or PrN-pYEE and IntN-SrcSH2, respectively; ΔF
Splicing activity of the designed intein
Figure 1. Design of split HIV-1 protease for generating active protease by PTS. (a) Schematic representation of the split HIV-1
protease conjugated with the DnaE split intein.(b) Amino acid sequences of PrN-intN, IntC-prC, and Rpep-intN. The sequences
derived from HIV-1 protease and DnaE intein are colored with green and yellow, respectively.

Figure 2. PTS reaction between Rpep-intN and IntC-prC. (a) SDS-PAGE (15% acrylamide) analysis of the time course of the PTS
reaction. (b) Kinetic analysis of producing IntC determined by densitometric analysis of the SDS-PAGE. All experiments were
performed in triplicate, and error bars represent standard deviation.
To check the splicing activity of the engineered Npu intein in
vitro, IntC-prC protein and Ppep-intN, not having the HIV-1
protease sequence, were reacted. IntC-prC (5.0 µM) was mixed
with Rpep-intN (25 µM) in a splicing buffer (pH 7.0) at 25°C,
incubated for zero hours, and analyzed by SDS-PAGE (Figure
2). Increasing the incubation period, IntC-prC, shown at around
50 kDa (calcd = 49.7 kDa), was gradually decreased. The new
proteins were detected at around 10 kDa and 40 kDa,
corresponding to the splicing product (calcd = 11.7 kDa) and
IntC fragment (calcd = 39.4 kDa), respectively. By quantifying
the bands corresponding to the IntC fragment, the kinetic
parameter was estimated as an observed rate constant (kobs) of
(1.3 ± 0.2) x 10-4 s-1. The activity of the engineered Npu DnaE
intein was weaker than those of the native form Npu DnaE
having N-terminal residues (1–4) of HIV-1 protease, was
examined to assess whether the splicing product can work as an
active enzyme. The protease activity was evaluated using a
fluorogenic substrate (Fsbst) having the sequence of
Abz-Thr-Ile-Nle-Phe(pNO2)-Gln-Arg-NH2
(Abz
=
2-aminobenzoyl).27,28 While the fluorescence of the Abz group
can be quenched by the nitrophenyl moiety via Förster
resonance energy transfer, it can increase by the proteolytic
cleavage between Nle and Phe(pNO2). The IntC-prC protein
(5.0 µM) was reacted with PrN-intN (25 µM) in the presence of
Fsbst (100 µM) in a splicing buffer (pH 7.0) at 25°C, and the
fluorescence spectra were measured (Figure 3A). The
fluorescence intensities at 420 nm were increased as a sigmoid
function of time (Figure 3B). This result suggests that the
product protein generated by PTS can readily form the active
structure and cleave the substrate peptide. Moreover, the
enzyme activity was clearly observedeven when incubated for
two hours. After incubation of eight hours, the fluorescence
signal at 420 nm was almost saturated, indicating that the
substrate was largely cleaved by the generated enzyme. This
kinetic behavior of generated HIV-1 protease shows good
correlation with that for the PTS reaction between Rpep-intN
peptide and IntC-prC. In contrast, when IntC-prC was
incubated alone, fluorescence was not changed even after eight
hours of incubation (Figure 3B). This finding suggests that the
precursor protein, IntC-prC, is completely inactive. Since the
N-terminal four amino acid residues are critical to form the
dimeric structure of the protease, N-terminal truncation of
HIV-1 protease by four amino acids completely abolished the
enzymatic activity
(k
= 3.5 x 10-3 s-1)11 and the M86 variant of Ssp DnaB
obs
intein,21 but it was similar or slightly higher than those of the
wt Ssp DnaB (kobs = 4.1 x 10-5 s-1)12 and Ssp DnaE (kobs = 8.0 x
10-5 s-1)11 inteins. Although the designed intein has moderate
activity, it can be applicable for developing the detection
system.
Generation of active HIV-1 protease by PTS
Next, splicing reaction between IntC-prC protein and PrN-intN,
Detection of phosphopeptide and SrcSH2 protein using
HIV-1 protease reconstitution system
Finally,
a signal transduction system that depends on
ligand-protein interactions has been constructed based on
generation of active HIV-1 protease by PTS. In this system,
IDNCL was employed to convert the interaction event between
a ligand and a target protein into the production of the peptide
having the PrN-intN sequence. To this aim, Src homology
domain 2 of c-Src (SrcSH2) and its ligand, pTyr-Glu-Glu-Ile
(pYEEI; pY = phosphorylated tyrosine) peptide, were used as a
model binding pair (Figure 4A).29-31 SH2 proteins can
recognize the phosphorylation sites with the specific peptide
sequences. SrcSH2 binds to the Ac-pYEEI-NH2 peptide with a
dissociation constant of Kd = 0.5 µM.31 It has been questioned
Figure 3. Generation of active HIV-1 protease by PTS
reaction between PrN-intN and IntC-prC. (a) Time course of
fluorescence emission spectra of Fsbst (100 µM) incubated
with PrN-intN (25 µM) and IntC-prC (5 µM) at 25°C. λex =
380 nm. (b) Fluorescence intensity at 420 nm as a function of
incubation time in the presence (open circle) and absence
(closed square) of PrN-intN. All experiments were performed
in triplicate, and error bars represent standard deviation.

Figure 4. Signal transduction depending on the interaction of pYEEI and SrcSH2. (a) Structures of PrN-pYEEI and IntN-SrcSH2,
and schematic representation of the signal transduction system. (b) Fluorescence change of Fsbst as a function of IDNCL reaction
time. (c) Fluorescence change of Fsbst as a function of PrN-pYEEI in the absence (closed circle) and presence (closed square) of
Ac-pYEEI-NH2 (50 µM). (d) Fluorescence change of Fsbst as a function of PrN-pYEE (open triangle) concentrations. All
experiments were performed in triplicate, and error bars represent standard deviation.
whether the input signals from the moderate binding pair can
be discriminated by IDNCL-PTS. The input peptide, named
PrN-pYEEI, which contains PrN and pYEEI sequences via a
thioester linkage, was designed and synthesized. Moreover, the
PrN-pYEE peptide, with a lack of the isoleucine residue of
PrN-pYEEI, was also synthesized to compare whether the
difference of binding affinities affects the output signal. The
input protein, named IntN-SrcSH2, which contains intN and
SrcSH2 sequences via a flexible (Gly-Gly-Ser)12 linker, was
also designed and expressed in E. coli. When the pYEEI
sequence of input peptide can interact with SrcSH2 of the input
protein, IDNCL occurs between the peptide and protein and
can provide the PrN-intN sequence at the N-terminal region of
SrcSH2.
IDNCL between PrN-pYEEI (10 µM) ligand and
IntN-SrcSH2 (4 µM) protein was performed for zero, two, four,
six, and eight hours of incubation. The reaction was directly
confirmed by RP-HPLC, but the product was almost
undetectable due to the low reaction yield (≤ 1%). After the
IDNCL reaction, IntC-prC (2.6 µM) and Fsbst (100 µM) were
added and incubated at room temperature for eight hours to
promote PTS at pH 7.4. To increase the enzyme activity, pH
was shifted at around 4.5,[26] and fluorescence spectra were
recorded before and after incubation at 25°C for several hours.
Figure 4B shows the fluorescence change at 420 nm as a
function of incubation time for IDNCL. Fluorescence changes
clearly appeared after the increased reaction time of IDNCL
between PrN-pYEEI and IntN-SrcSH2. Thus, the interaction
events between pYEEI and SrcSH2 were successfully
transduced into the protease activity. Under the eight hours
incubation condition, dose dependency of PrN-pYEEI was
examined. Figure 4C shows the IDNCL-PTS results using
various concentrations of PrN-pYEEI (0, 1.25, 2.5, 5, and 10
µM) and IntN-SrcSH2 (4 µM). The fluorescence changes
reflecting the enzyme activity were dose-dependently increased.
When assuming that the fluorescence change is correlated with
the complex formation between pYEEI and SrcSH2, the
binding affinity of PrN-pYEEI for IntN-SrcSH2 was estimated
as Kd = 1.1 µM, which is a similar value of the reported one
(0.5 µM). Moreover, the weak interaction between PrN-pYEE
and IntN-SrcSH2 was also assessed by IDNCL-PTS, and the
affinity was estimated as Kd = 28 µM (Figure 4D). In contrast,
incubation of PrN-pYEEI (10 µM) with IntN-SrcSH2 (4 µM)
in the presence of Ac-pYEEI-NH2 (50 µM) as a competitor
caused a slight increase in fluorescence (Figure 4C). These
findings indicate that the specific interaction between pYEEI
peptide and SrcSH2 promotes an NCL reaction for generating
the PrN-intN sequence at the N-terminal site upon SrcSH2.
Thus, binding events between a ligand and a protein, such as
pYEEI and SrcSH2, were successfully transduced into the
protease activity. This can be used for generating a variety of
functional proteins, such as several enzymes and recognition
modules.
4. Conclusion
In this study, an artificial signal transduction system based on
IDNCL-PTS has been successfully developed using an
engineered split intein, derived from Npu DnaE, and HIV-1
protease. Truncation of only four amino acid residues at
N-terminus of HIV-1 protease completely abolished the
protease activity due to lack of the active dimer structure. The
attachment of the N-terminal residues into the truncated HIV-1
protease by PTS using the engineered Npu DnaE intein
generated the active enzyme that cleaves the fluorogenic
substrate. By combining IDNCL with the PTS reaction, the
signal transduction system was successfully constructed. The
protease activity was clearly observed depending on the
interactions of input PrN-pYEEI and PrN-pYEE with input
IntN-SrcSH2 protein. The HIV-1 protease activity generated by
IDNCL-PTS might activate several enzymes working in
naturally occurring signal cascades. Indeed, the low reaction
yield of IDNCL has to be improved by changing the chemical

structure of the thioester linkage to increase the detection limit.
While the improvement of the IDNCL efficiency is necessary,
the artificial signal transduction system might be available in
synthetic biology in the future.
28. T. Takahashi, D. Yana, H. Mihara, ChemBioChem 2006,
7, 729.
29. G. Waksman, S. E. Shoelson, N. Pant, D. Cowburn, J.
Kuriyan, Cell 1993, 72, 779.
30. N. J. de Mol, F. J. Dekker, I. Broutin, M. J. E. Fischer, R.
M. J. Liskamp, J. Med. Chem. 2005, 48, 753.
31. P. S. Charifson, L. M. Shewchuk, W. Rocque, C. W.
Hummel, S. R. Jordan, Ch. Mohr, G. J. Pacofsky, M. R.
Peel, M. Rodriguez, D. D. Sternbach, T. G. Consler,
Biochemistry 1997, 36, 6283.
Acknowledgement
This work was supported by JSPS KAKENHI Grand Number
15K01818.
References
1.
2.
J. S. Sebolt-Leopard, R. Herrera, Nat. Rev. 2004, 4, 937.
S. H. Lecker, A. L. Goldberg, W. E. Mitch, J. Am. Soc.
Nephrol. 2006, 17, 1807.
3.
4.
5.
6.
T. Kaufmann, A. Strasser, P. J. Jost, Cell Death Differ.
2012, 19, 42.
A. Ashkenazi, V. M. Dixit, Curr. Opin. Cell Biol. 1999,
11, 255.
A. Galarneau, M. Primeau, L.-E. Trudeau, S. W.
Michnick, Nat. Biotechnol. 2002, 20, 619.
S. Cabantous, H. B. Nguyen, J.-D. Pedelacq, F. Koraïchi,
A. Chaudhary, K. Ganguly, M. A. Lockard, G. Favre, T.
C. Terwilliger, G. S. Waldo, Sci. Rep. 2013, 3, 2854.
S. Sakamoto, K. Kudo, J. Am. Chem. Soc. 2008, 130,
9574.
A. S. Dixon, M. K. Schwinn, M. P. Hall, K. Zimmerman,
P. Otto, T. H. Lubben, B. L. Butler, B. F. Binkowski, T.
Machleidt, T. A. Kirkland, M. G. Wood, C. T. Eggers, L.
P. Encell, K. V. Wood, ACS Chem. Biol. 2016, 11, 400.
M. C. Wehr, R. Laage, U. Bolz, T. M. Fischer, S.
Grünewald, S. Scheek, A. Bach, K.-A. Nave, M. J.
Rossner, Nat. Methods 2006, 3, 985.
7.
8.
9.
10. H. Wu, Z. Hu, X.-Q. Liu, Proc. Natl. Acad. Sci. USA
1998, 95, 9226.
11. J. Zettler, V. Schütz, H. D. Mootz, FEBS Lett. 2009, 583,
909.
12. C. Ludwig, M. Pfeiff, U. Linne, H. D. Mootz, Angew.
Chem. Int. Ed. 2006, 45, 5218.
13. H. Wu, M.-Q. Xu, X.-Q. Liu, Biochim. Biophys. Acta
1998, 1387, 422.
14. J. Shi, T. W. Muir, J. Am. Chem. Soc. 2005, 127, 6198.
15. G. P. Liszczak, Z. Z. Brown, S. H. Kim, R. C. Oslund, Y.
David, T. W. Muir, Proc. Natl. Acad. Sci. USA 2017, 114,
681.
16. A. Kanno, T. Ozawa, Y. Umezawa, Anal. Chem. 2006, 78,
556.
17. H. D. Mootz, E. S. Blum, A. B. Tyszkiewicz, T. W. Muir,
J. Am. Chem. Soc. 2003, 125, 10561.
18. E. C. Schwartz, L. Saez, M. W. Young, T. W. Muir, Nat.
Chem. Biol. 2007, 3, 50.
19. T. Sonntag, H. D. Mootz, Mol. Biosyst. 2011, 7, 2031.
20. T. Takahashi, A. Saito, ChemistrySelect 2016, 1, 1768.
21. J. H. Appleby-Tagoe, I. V. Thiel, Y. Wang, Y. Wang, H.
D. Mootz, X.-Q. Liu, J. Biol. Chem. 2011, 286, 34440.
22. T. D. Meek, B. D. Dayton, B. W. Metcalf, G. B. Dreyer, J.
E. Strickler, J. G. Gorniak, M. Rosenberg, M. L. Moore,
V. W. Magaard, C. Debouck, Proc. Natl. Acad. Sci. USA
1989, 86, 1841.
23. I. T. Weber, M. Miller, M. Jaskolski, J. Leis, A. M.
Skalka, A. Wlodawer, Science 1989, 243, 928.
24. S. Choudhury, L. Everitt, S. C. Pettit, A. H. Kaplan,
Virology 2003, 307, 204.
25. S. Ellilä, J. M. Jurvansuu, H. Iwaï, FEBS Lett. 2011, 585,
3471.
26. A. S. Aranko, J. S. Oeeming, D. Zhou, T. Kajander, A.
Wlodawer, H. Iwaï, Mol. Biosyst. 2014, 10, 1023.
27. Z. Szeltner, L. Polgár, J. Biol. Chem. 1996, 271, 32180.