A novel N-terminal degradation reaction of peptides via N-amidination

Article abstract of DOI:10.1016/j.bmcl.2016.02.058

The cleavage of amide bonds requires considerable energy. It is difficult to cleave the amide bonds in peptides at room temperature, whereas ester bonds are cleaved easily. If peptide bonds can be selectively cleaved at room temperature, it will become a powerful tool for life science research, peptide prodrug, and tissue-targeting drug delivery systems. To cleave a specific amide bond at room temperature, the decomposition reaction of arginine methyl ester was investigated. Arginine methyl ester forms a dimer; the dimer releases a heterocyclic compound and ornithine methyl ester at room temperature. We designed and synthesized N-amidinopeptides based on the decomposition reaction of arginine methyl ester. Alanyl-alanine anilide was used as the model peptide and could be converted into N-degraded peptide, alanine anilide, via an N-amidination reaction at close to room temperature. Although the cleavage rate in pH 7.4 phosphate buffered saline (PBS) at 37?°C was slow (t_1/2?=?35.7?h), a rapid cleavage rate was observed in 2% NaOH aq (t_1/2?=?1.5?min). To evaluate the versatility of this reaction, a series of peptides with Lys, Glu, Ser, Cys, Tyr, Val, and Pro residue at the N-terminal were synthesized; they showed rapid cleavage rates of t_1/2values from 1?min to 10?min.

Full text of DOI:10.1016/j.bmcl.2016.02.058

Bioorganic & Medicinal Chemistry Letters 26 (2016) 1690–1695
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
A novel N-terminal degradation reaction of peptides
via N-amidination
⇑
Yoshio Hamada
Medicinal Chemistry Laboratory, Kobe Pharmaceutical University, Motoyamakita, Higashinada, Kobe 658-8558, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The cleavage of amide bonds requires considerable energy. It is difficult to cleave the amide bonds in pep-
tides at room temperature, whereas ester bonds are cleaved easily. If peptide bonds can be selectively
cleaved at room temperature, it will become a powerful tool for life science research, peptide prodrug,
and tissue-targeting drug delivery systems. To cleave a specific amide bond at room temperature, the
decomposition reaction of arginine methyl ester was investigated. Arginine methyl ester forms a dimer;
the dimer releases a heterocyclic compound and ornithine methyl ester at room temperature. We
designed and synthesized N-amidinopeptides based on the decomposition reaction of arginine methyl
ester. Alanyl-alanine anilide was used as the model peptide and could be converted into N-degraded pep-
tide, alanine anilide, via an N-amidination reaction at close to room temperature. Although the cleavage
rate in pH 7.4 phosphate buffered saline (PBS) at 37 °C was slow (t1/2 = 35.7 h), a rapid cleavage rate was
observed in 2% NaOH aq (t1/2 = 1.5 min). To evaluate the versatility of this reaction, a series of peptides
with Lys, Glu, Ser, Cys, Tyr, Val, and Pro residue at the N-terminal were synthesized; they showed rapid
cleavage rates of t1/2 values from 1 min to 10 min.
Received 25 December 2015
Revised 5 February 2016
Accepted 19 February 2016
Available online 21 February 2016
Keywords:
Amide bond cleavage
Edman degradation
N-amidination
N-Amidinopeptide
N-terminal degradation
Ó 2016 Elsevier Ltd. All rights reserved.
The selective cleavage of a specific peptide bond at room tem-
perature will be a powerful tool for life science research, peptide
prodrugs, and tissue-targeting drug delivery systems. However,
the cleavage of amide bonds requires considerable energy,1
whereas ester bonds can be easily cleaved. Many ester-type pro-
drugs that release the corresponding parent drugs with a hydroxy
group under physiological conditions have been reported.2,3 Previ-
ously, we reported a series of novel ester-type prodrugs that could
be converted to the parent drugs under physiological conditions
without using enzymes.4–8 Although an amide bond is difficult to
cleave at room temperature, selective amide bond cleavage reac-
tions are known in nature. For example, a protein, intein, under-
goes protein splicing that involves a particular peptide bond
cleavage.9,10 To develop a reaction for specific amide bond cleav-
age, we investigated the decomposition reaction of arginine methyl
ester reported by Photaki et al.,11 as shown in Figure 1A. In this
reaction, the guanidino group of an arginine methyl ester attacks
the ester carbonyl carbon of another arginine methyl ester, forming
an arginine dimer. Next, the N-terminal amino group of the dimer
attacks the guanidine carbon within the molecule, forming a hete-
rocyclic compound and ornithine methyl ester. Surprisingly, this
decomposition reaction proceeds at room temperature under neu-
tral conditions. This N-aminoacyl guanidine structure of the dimer
releases an N-terminal amino acid moiety at room temperature as
shown in Figure 1B. However, the guanidine moiety does not con-
tain an amide bond; this compound is also not a common peptide.
We assumed that the driving force of this decomposition reac-
tion is the release of the heterocyclic compound that is stabilized
because of a conjugate structure. Because the N-amidino aminoa-
cyl structure in which the order of the acyl and guanidino groups
is reversed is thought to release the same heterocyclic compound
to the N-aminoacyl guanidine compound (Fig. 1B), we envisioned
that the release of the heterocyclic compound would be the driving
force for the cleavage reaction of an amide bond. The N-amidina-
tion of a peptide can be applied to the N-terminal degradation
reaction of the peptide at room temperature. Hence, we designed
a novel N-terminal cleavage reaction of peptides using this reac-
tion and synthesized a series of N-amidinopeptides 1–8 and 18.
Peptides 1–8 were synthesized using the common solution-
phase peptide synthesis method as shown in Schemes 1 and 2.
The peptide bond formation was performed using 1-ethyl-3-(3-
dimethylaminopropyl)carbodiimide HCl (EDC) in the presence of
1-hydroxybenzotriazole (HOBt) as the coupling reagent. The amid-
ination reactions for preparing 16a–g were performed using
N,N0-bis(tert-butoxycarbonyl)-1H-pyrazole-1-carboxamidine
in
acetonitrile or ethyl acetate. The Boc-deprotections of 9 and 11
were performed using anisole and 4 N HCl/dioxane. The
⇑
Tel.: +81 78 441 7555.
E-mail address: pynden@gmail.com
http://dx.doi.org/10.1016/j.bmcl.2016.02.058
0960-894X/Ó 2016 Elsevier Ltd. All rights reserved.

Y. Hamada / Bioorg. Med. Chem. Lett. 26 (2016) 1690–1695
1691
NH
NH
NH
A
B
NH
NH
NH
NH
O
O
H2N
H2N
H2N
H2N
NH2
NH
NH2
H
N
O
NH
NH
H2N
H2N
O
HN
HN
O
O
H2N
H2N
O
O
NH
O
O
heterocyclic
compound
ornithine
methyl ester
arginine methyl ester
R2
arginine dimer
R2
R2
O
NH
O
H
H2N
H2N
N
N
H
N
H2N
N
H
R1
R1
H
O
R1
O
NH
O
R1
O
O
HN
HN
N-aminoacyl guanidine compound
N-amidinopeptide
NH
NH
HN
HN
novel N-terminal
degradation reaction
amidination
R2
O
H2N
N
H
R1
O
Figure 1. (A) Decomposition reaction of arginine methyl ester. (B) Design of an N-terminal degradation reaction of peptides via N-amidinopeptides.
H
N
H
N
a
b
Boc
Boc
OH
N
H
H2N
N
H
O
O
O
9
10
Boc Ala OH
O
O
H
N
H
N
H
c
b
H2N
N
Boc
Boc
N
H
N
H
O
O
11
12
O
O
H
N
H
N
H
H
N
H
N
e
d
N
H2N
N
H
N
H
N
O
NH
O
Boc
13
1
Scheme 1. Reagents and conditions: (a) aniline, EDCꢀHCl, HOBt, acetonitrile, 4 h; (b) anisole 4 N HCl/dioxane, 2 h; (c) Boc-Ala-OH, EDCꢀHCl, HOBt, acetonitrile, 4 h; (d) N,N0-bis
(tert-butoxycarbonyl)-1H-pyrazole-1-carboxamidine, ethyl acetate, 1 day; (e) TFA, 2 h.
Fmoc-deprotections of 14a–g were performed using 20% diethy-
lamine in dioxane. The final deprotections for preparing of 1–8
were performed using trifluoroacetic acid (TFA) with/without
cation scavengers, m-cresol, and ethyl mercaptan. Because the
deprotection reaction for preparing 5 afforded some by-products
with an S-trityl group or an S-tert-butyl group, the trityl group of
16d was deprotected using iodide in ethyl acetate before the TFA
treatment. Peptides 1–8 were obtained as the TFA salts and stored
in a refrigerator. Peptides 17 and 18 were synthesized via 9-fluo-
renylmethoxycarbonyl (Fmoc)-based solid-phase peptide synthe-
sis using a 2-chlorotritylchloride resin as shown in Scheme 3,
and the details are given in the Supporting information. The Fmoc
group was removed with 20% piperidine in dimethylformamide
(DMF). Peptide bonds were formed using diisopropylcarbodiimide
(DIPCDI) in the presence of HOBt as the coupling reagent. N-amid-
ination of peptides was achieved using N,N0-bis(tert-butoxycar-
bonyl)-1H-pyrazole-1-carboxamidine in DMF. After the peptide
chains were elongated to their desired length, they were cleaved
from the resin using TFA in the presence of m-cresol, thioanisole,
and water. The peptides were all purified by preparative reverse
phase high-performance liquid chromatography (RP-HPLC).
To evaluate our strategy for developing the N-terminal degrada-
tion reaction, we synthesized peptide 1, in which the N-terminal
amino group of the alanyl-alanine anilide 12 was used as the
model peptide and converted to the guanidino group. First, peptide
1 was incubated with pH 7.4 phosphate buffered saline (PBS) buf-
fer under physiological conditions at 37 °C and evaluated by high-
performance liquid chromatography (HPLC) using a reverse-phase
C18 column and a linear gradient system of acetonitrile and 0.1%
aqueous TFA. Unfortunately, the cleavage rate of the N-terminal

1692
Y. Hamada / Bioorg. Med. Chem. Lett. 26 (2016) 1690–1695
H
H
N
a
b
N
Fmoc Xaa OH
Fmoc Xaa
N
H
H
Xaa
N
H
O
O
14a
14b
14c
15a
15b
15c
(Xaa = -Lys(Boc)-)
(Xaa = -Lys(Boc)-)
t
t
(Xaa = -Glu(OBu )-)
(Xaa = -Glu(OBu )-)
t
t
(Xaa = -Ser(Bu )-)
(Xaa = -Ser(Bu )-)
14d (Xaa = -Cys(trityl)-)
14e (Xaa = -Tyr(But)-)
14f (Xaa = -Val-)
15d (Xaa = -Cys(trityl)-)
15e (Xaa = -Tyr(But)-)
15f (Xaa = -Val-)
14g (Xaa = -Pro-)
15g (Xaa = -Pro-)
H
N
H
N
Boc NH
H2N
HN
c
d or e,d
Xaa
N
H
Xaa
N
H
Boc
N
O
O
16a (Xaa = -Lys(Boc)-)
2 (Xaa = -Lys-)
3 (Xaa = -Glu-)
16b (Xaa = -Glu(OBut)-)
t
16c
16d
16e
(Xaa = -Ser(Bu )-)
4
(Xaa = -Ser-)
(Xaa = -Cys(trityl)-)
5 (Xaa = -Cys-)
6 (Xaa = -Tyr-)
t
(Xaa = -Tyr(Bu )-)
7
(Xaa = -Val-)
16f (Xaa = -Val-)
16g (Xaa = -Pro-)
8 (Xaa = -Pro-)
Scheme 2. Reagents and conditions: (a) 10, EDCꢀHCl, HOBt, DMF, 4 h; (b) 20% diethylamine, dioxane, 1 h; (c) N,N0-bis(tert-butoxycarbonyl)-1H-pyrazole-1-carboxamidine,
acetonitrile, 1 day; (d) m-cresol, ethyl mercaptan, TFA, 2 h; (e) I2, ethyl acetate.
But OBut
Trityl
OBut
OBut
(a,b) x 8
Fmoc Phe
Fmoc Ser Glu Val Asn Pns Asp Ala Glu Phe
c
a,b,d
Trityl
H
Ser Glu Val Asn Pns Asp Ala Glu Phe OH
17
But OBut
OBut
OBut
Boc NH
Ala Ser Glu Val Asn Pns Asp Ala Glu Phe
Boc
N
= 2-chlorotrityl resin
c
Pns = (2R,3S)-3-amino-2-
H2N
HN
hydroxy-4-phenyl
butyric acid
Ala Ser Glu Val Asn Pns Asp Ala Glu Phe OH
18
Scheme 3. Reagents and conditions: (a) Fmoc-protected amino acid, HOBt, DIPCDI, DMF, 2 h; (b) 20% piperidine, DMF, 20 min; (c) thioanisole, m-cresol, H2O, TFA; (d) N,N0-bis
(tert-butoxycarbonyl)-1H-pyrazole-1-carboxamidine, DMF, 1 day.
Figure 2. (A) Time course of peptide 1 in incubation mediums at 37 °C. (B) The correlation between the N-degradation reaction of peptide 1 and the respective pH values of
the incubation media. The k values are the rate constants of the N-terminal degradation reaction. (C) HPLC profiles of peptide 1 in 2% NaOH aq at 37 °C.

Y. Hamada / Bioorg. Med. Chem. Lett. 26 (2016) 1690–1695
1693
Figure 3. (A) HPLC profiles of peptides 2–4 and 6–8. (B) HPLC profiles of peptides 2–4 and 6–8 after the incubation in 2% NaOH aq for 5 min at 37 °C. (C) Time course of
peptides 1–4 and 6–8 in 2% NaOH aq at 37 °C.
alanine residue was very slow in pH 7.4 PBS at 37 °C, as shown in
Figure 2A (t1/2 value: 35.7 h). This is the first N-terminal degrada-
tion reaction of peptides at room temperature. The N-terminal
cleavage reaction, Edman degradation,12,13 requires acidic and
respectively. Higher pH values appeared to accelerate the N-degra-
dation reaction, as shown in Figure 2B. To investigate the pH
dependency of the N-terminal degradation reaction, the rate con-
stants of peptide 1 in the respective incubation media were calcu-
lated using the fitting Eq. 1
heating conditions after the treatment of
a peptide with
phenylisothiocyanate. Next, the cleavage rates of peptide 1 in alka-
line media, with pH values ranging from 8.0 to 13.7 were mea-
sured. The cleavage rates (given as t1/2 values) in 2% aq NaHCO3
(pH = 8.9), 2% aq K2CO3 (pH = 11), 0.5% aq NaOH (pH = 13) and 2%
aq NaOH (pH = 13.7) were 8.1 h, 3.4 h, 5.8 min, and 1.5 min,
½Aꢁt ¼ AMAX ꢂ ð1 ꢃ Expðꢃk ꢂ tÞÞ
ð1Þ
where, t is the incubation time; k is the rate constant of the N-ter-
minal degradation reaction; AMAX is the maximum concentration of
alanine anilide; and [A] is the concentration of alanine anilide.
Figure 4. (A) HPLC profile of peptide 5 in 2% NaOH aq at 37 °C. (B) Time course of peptide 5 in 2% NaOH aq at 37 °C.

1694
Y. Hamada / Bioorg. Med. Chem. Lett. 26 (2016) 1690–1695
A
bation in 2% NaOH aq showed the most rapid cleavage rate, as
O
shown in Figure 2A. This peptide is stable as a TFA salt and released
the N-terminal-degraded peptide, alanine anilide 10, time-depen-
dently without the formation of any by-product as shown in
Figure 2C.
H
N
H
N
H2N
N
H
NH
O
B
HS
To evaluate the versatility of the N-terminal degradation reac-
tion, peptides 2–8 possessing Lys, Glu, Ser, Cys, Tyr, Val, and Pro
residues at the N-terminal position were synthesized. Because
these peptides contain a basic group, an acidic group, an OH group,
an SH group, an aromatic ring, a sterically hindered branched alkyl
group, and a cyclic structure, respectively, on the side chains, they
cover all types of amino acid residues for evaluating the N-terminal
degradation reaction. Peptides 2–8 were incubated in 2% NaOH aq
at 37 °C and measured by HPLC. Peptides 2–4 and 6–8 rapidly
released the N-terminal-degraded peptide, alanine anilide, without
any by-product as shown in Figure 3A and B. Although the hetero-
cyclic compounds that were released from most of the
N-amidinopeptides had a high hydrophilicity and low UV absorp-
tion coefficient and could not be identified by the HPLC analysis
using an ODS column and a UV detector. Although the heterocyclic
compound that was released from peptide 6 with a Tyr residue
could be identified by HPLC, as shown in Figure 3B, all products
released from N-amidinopeptides 1–8 were identified using ESI-
Mass analysis (see Supporting information). In the case of peptide
5 with a Cys residue, an unidentified peak appeared on the HPLC
trace at ꢄ20.2 min after the incubation in 2% NaOH aq at 37 °C
as shown in Figure 4A. The peak at ꢄ20.2 min immediately
appeared on the HPLC trace after the incubation and then disap-
peared slowly. Because the peaks of the unidentified compound
and peptide 5 showed an almost equal rate (55:45, Fig. 4B), they
may be in equilibrium. There are two possibilities for the uniden-
tified compound—intermediate A and B as shown in Figure 5. A
peak that showed the same molecular weight to the intermediate
A
B
2% NaOH
O
H
N
OH
H
HN
N
H2N
H2N
Ala NH-Ph
Ala NH-Ph
HN
HS
S
intermediate A
intermediate B
H
N
H
HN
N
O
H2N
HN
HS
O
alanine anilide
Figure 5. N-terminal degradation reaction and intermediates of peptide 5 with a
Cys residue.
The pH values (ꢃlog[H+]) of the respective incubation media
showed
a
negative correlation (coefficient of correlation,
r = –0.9814) with the –log k values (k, rate constant) of the
N-terminal degradation reaction, as shown in Fig. 2B—which
means that the rate constants of the N-terminal degradation reac-
tion appeared to be positively correlated to the concentration of
the hydroxide ions [OHꢃ] in the incubation media. Thus, the incu-
Figure 6. (A) HPLC profiles of the crude peptides, 17 and 18, after their cleavage from the resin. (B) HPLC profiles of the preparative RP-HPLC purified peptides, 17 and 18.

Y. Hamada / Bioorg. Med. Chem. Lett. 26 (2016) 1690–1695
1695
Figure 7. (A) HPLC profile of N-amidino-decapeptide 18 in 2% aq NaOH at 37 °C. (B) Degradation reaction progress of N-amidino-decapeptide 18 in 2% aq NaOH at 37 °C.
A and B was identified on the ESI-Mass trace obtained from the
incubated solution of peptide 5. Intermediate A, which was
immediately produced before the N-terminal cleavage reaction, is
unlikely, because such a peak did not appear on the HPLC traces
of other N-amidinopeptides. As peptide 5 with a Cys residue can
only form an intermediate with a five-membered ring, thiazolidine,
the unidentified peak at ꢄ20.2 min seem to be intermediate B. The
t1/2 value of peptide 5 in 2% NaOH aq at 37 °C was 3.4 min, given
the existence of intermediate B. As the cleavage of peptide 7
(t1/2 = 9.2 min) and 8 (t1/2 = 7.4 min) showed slower cleavage rates
than peptides 1–6 with a t1/2 value from 1.2 min to 3.4 min as
shown in Figure 3C. The sterically hindered group and ring struc-
ture of Val and Pro residues, respectively, may prevent the forma-
tion of a five-membered ring transition state.
degradation reaction proceeds in neutral-to-weak alkaline media,
whereas Edman degradation requires acidic and heating condi-
tions. Because our degradation proceeds under different reaction
conditions than Edman degradation, it has great potential to
become an important tool for life science research.
Acknowledgments
This study was supported in part by the Grants-in-Aid for Scien-
tific Research (C) from MEXT (Ministry of Education, Culture,
Sports, Science and Technology), Japan (KAKENHI Nos. 23590137
and 26460163). We thank Dr. O. Miyata and Dr. A. Takeuchi for
their help in preparing the manuscript and measuring ESI-Mass
spectra, respectively.
We applied the N-terminal degradation reaction to a long-chain
peptide. Peptide 17 contains an amino acid sequence from our pre-
viously reported b-secretase inhibitory peptide and an unusual
amino acid, (2R,3S)-3-amino-2-hydroxy-4phenylbutyric acid
(Pns), as a transition-state analog.14–19 We designed and synthe-
sized N-amidino-decapeptide 18 based on peptide 17. The HPLC
profiles of 18, before and after purification via preparative RP-
HPLC, are shown in Figure 6. N-Amidino-decapeptide 18 was incu-
bated in 2% aq NaOH at 37 °C and analyzed by HPLC. As shown in
Figure 7, peptide 18 released the N-terminal-degraded peptide 17
in a time-dependent manner, and it was identified using HPLC
retention time and ESI-Mass experiment. Peptide 18 released 17
rapidly with a t1/2 value of 1.7 min and this is close to that of pep-
tide 1 (t1/2 = 1.5 min), which has the same N-terminal amino acid
residue, Ala. These results indicated that the length of a peptide
chain does not affect the N-terminal degradation reaction.
In conclusion, we designed a novel N-terminal degradation
reaction and synthesized a series of N-amidinopeptides, in which
the order of the acyl and guanidino groups of N-aminoacyl
guanidine structure in the dimer of arginine methyl ester was
reversed. The driving force of this reaction may be the release of
a heterocyclic compound that is stabilized because of a conjugate
structure. A specific amide bond, an N-terminal amino acid residue
could be successfully cleaved at almost room temperature with
rapid rates of t1/2 values from 1 min to 10 min. Moreover, we
applied the N-terminal degradation reaction to a long-chain
peptide. N-Amidino-decapeptide 18 showed a rapid N-terminal
degradation rate, with a t1/2 value of 1.7 min. Our degradation reac-
tion provided the same product as the Edman degradation. This
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2016.02.
058.
References and notes
1. Simplício, A. L.; Clancy, J. M.; Gilmer, J. F. Molecules 2008, 13, 519.
2. Stella, V. J.; Nti-Addae, K. W. Adv. Drug Deliv. Rev. 2007, 59, 677.
3. Gomes, P.; Vale, N.; Moreira, R. Molecules 2007, 12, 2484.
4. Hamada, Y.; Ohtake, J.; Sohma, Y.; Kimura, T.; Hayashi, Y.; Kiso, Y. Bioorg. Med.
Chem. 2002, 10, 4155.
5. Hamada, Y.; Matsumoto, H.; Kimura, T.; Hayashi, Y.; Kiso, Y. Bioorg. Med. Chem.
Lett. 2003, 13, 2727.
6. Hamada, Y.; Matsumoto, H.; Yamaguchi, S.; Kimura, T.; Hayashi, Y.; Kiso, Y.
Bioorg. Med. Chem. 2004, 12, 159.
7. Hayashi, Y.; Skwarczynski, M.; Hamada, Y.; Sohma, Y.; Kimura, T.; Kiso, Y. J.
Med. Chem. 2003, 46, 3782.
8. Skwarczynski, M.; Sohma, Y.; Noguchi, M.; Kimura, M.; Hayashi, Y.; Hamada,
Y.; Kimura, T.; Kiso, Y. J. Med. Chem. 2005, 48, 2655.
9. J Peter Gogarten, J. P.; Hilario, E. BMC Evolutionary Biol. 2006, 6, 94.
10. Yasuhiro Anraku, Y.; Mizutani, R.; Satow, Y. IUBMB Life 2005, 57, 563.
11. Photaki, I.; Yiotakis, A. J. Chem. Soc., Perkin Trans. 1 1976, 24, 259.
12. Edman, P. Acta Chem. Scand. 1950, 1, 283.
13. Edman, P. Begg. G. Eur J Biochem. 1967, 1, 80.
14. Hamada, Y.; Kiso, Y. Expert Opin. Drug Discov. 2009, 4, 391.
15. Hamada, Y.; Kiso, Y. Expert Opin. Drug Discov. 2012, 7, 903.
16. Hamada, Y.; Kiso, Y. Expert Opin. Drug Discov. 2013, 8, 709.
17. Hamada, Y. SOJ Pharm. Pharm. Sci. 2014, 1, 1.
18. Hamada, Y.; Kiso, Y. In Amino Acids, Peptides and Proteins; Royal Society of
Chemistry: London, 2015; Vol. volume 39, 114.
19. Hamada, Y.; Ishiura, S.; Kiso, Y. ACS Med. Chem. Lett. 2012, 3, 193.