Development of an efficient liquid-phase peptide synthesis protocol using a novel fluorene-derived anchor support compound with Fmoc chemistry; AJIPHASE

Article abstract of DOI:10.1016/j.tetlet.2012.02.006

The development of a novel fluorene-type anchor support molecule and a liquid-phase peptide synthesis protocol (AJIPHASE) is reported. With this support molecule, the synthesis of a protected peptide with a free carboxyl group could be carried out by repe

Full text of DOI:10.1016/j.tetlet.2012.02.006

Tetrahedron Letters 53 (2012) 1936–1939
Contents lists available at SciVerse ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Development of an efficient liquid-phase peptide synthesis protocol
using a novel fluorene-derived anchor support compound with Fmoc
chemistry; AJIPHASEÒ
Daisuke Takahashi a, , Takashi Yamamoto b
⇑
a Research Institute for Bioscience Products and Fine Chemicals, Ajinomoto Co., Inc. 1730 Hinaga Yokkaich, Mie 510-0885, Japan
b Exploratory Research Laboratories, Ajinomoto Pharmaceuticals Co., Ltd, 1-1, Suzuki-cho, Kawasaki-ku, Kawasaki-shi, Kanagawa 210-8681, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The development of a novel fluorene-type anchor support molecule and a liquid-phase peptide synthesis
protocol (AJIPHASEÒ) is reported. With this support molecule, the synthesis of a protected peptide with a
free carboxyl group could be carried out by repeated coupling/deprotection reactions and isolation by
simple precipitation. Cleavage is performed under mild acidic conditions (2% TFA/CHCl3 or hexafluoroiso-
propanol) to release the products in good yield and purity. There were no signs of either diketopiperazine
formation or the removal of protective groups in the side-chain.
Received 21 December 2011
Revised 30 January 2012
Accepted 2 February 2012
Available online 9 February 2012
Keywords:
Ó 2012 Elsevier Ltd. All rights reserved.
Liquid-phase peptide synthesis
Fluorene-derived anchor support molecule
AJIPHASEÒ
The market for peptide-based pharmaceuticals is growing rap-
idly due to an increased number of successful therapeutic targets
and improved delivery methodologies. Furthermore, more than
100 peptide drug candidates are available in the clinic and a large
number of pre-clinical development projects are ongoing world-
wide. Thus, there is a growing need for a generally applicable pep-
tide synthesis protocol, which could be used to prepare peptides
with high purity and at low cost. The solid-phase peptide synthesis
(SPPS) method has been considered the gold standard for decades.
While, in solution phase, peptide coupling reactions can be carried
out under homogenous conditions using lesser amounts of amino
acids and coupling reagents compared to SPPS. Therefore, peptide
synthesis methods in solution phase offer economic and qualita-
tive advantages in process development for large-scale synthesis.
However, conventional solution-phase peptide synthesis also has
several drawbacks: process development is time-consuming,
workup and reaction conditions must be optimized for intermedi-
ate peptides, and it can be difficult to synthesize peptides with
hydrophilic, hydrophobic, or long chains.
polyethylene glycol,2 low molecular weight polystyrene,3 and
fluorous materials4 have been used as support molecules at the
C-terminus. However, there has been no report on their application
for industrial-scale synthesis, presumably due to the inefficient
removal of residual reagents and the difficult in handling of the
support compounds. Using benzyl alcohol derivative 1, which has
three long alkoxy chains, Tamaki reported a protocol in which
reaction intermediates were purified by precipitation combined
with flash chromatography (Fig. 1).5 Chiba6 and Sunazuka7 used
this protocol with compound 1 or its derivatives to purify synthe-
sized peptides by precipitation and/or extraction. With this sup-
port compound 1, the solubility of the peptide under the
coupling reaction conditions was significantly improved to achieve
sufficient reactivity, and the intermediate peptide could be easily
isolated by simple precipitation using polar organic solvents. Thus,
this LPPS protocol with compound 1 follows the simple ‘reaction
and washing’ concept of SPPS.
The support molecules that have seen practical use to date have
mostly been limited to benzyl alcohol derivatives, like compound
1, and they are associated with a risk of diketopiperazine formation
To overcome these problems, several solution-phase peptide
synthesis1 and liquid-phase peptide synthesis (LPPS) protocols
have been developed and reported to date. The LPPS is the methods
using soluble support molecules instead of insoluble resins in SPPS,
therefore possessing the combined advantages of both SPPS and
solution-phase peptide synthesis. In previously reported LPPS,
OC18H37
HO
OC18H37
OC18H37
1
⇑
Corresponding author. Tel.: +81 59 346 0122; fax: +81 59 346 0127.
E-mail address: daisuke_takahashi@ajinomoto.com (D. Takahashi).
Figure 1. Support compound 1 for liquid-phase peptide synthesis.
0040-4039/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2012.02.006

D. Takahashi, T. Yamamoto / Tetrahedron Letters 53 (2012) 1936–1939
1937
during peptide elongation, based on our examination as described
below. In this Letter, we describe the development of an Fmoc
chemistry-based novel LPPS protocol that uses fluorene derivatives
possessing long-chain alkoxy groups as ‘anchor’ support molecules,
and that shows no signs of diketopiperazine formation. In addition,
with a fluorene-derived support molecule, the peptide could be
cleaved off with the protective groups intact under weakly acidic
conditions. Therefore, this LPPS method is useful for the prepara-
tion of protected peptide acids for further fragment condensation.
We first used the support compound 1 for LPPS to synthesize an
Fmoc-protected peptide Fmoc-Tyr-Phe-Ser-Ala-Pro-Gly-OH. The
synthesis involves four steps. N-Fmoc-protected amino acid was
coupled to the anchor molecule, and this was followed by precip-
itation and washing using methanol. The Fmoc group was depro-
tected in the presence of Et2NH/DBU in chloroform. The resulting
benzyl ester of the N-terminal free amino acid was precipitated
and isolated by treatment with acetonitrile. The long alkoxy chains
of the anchor support compound 1 are the key structures in this
protocol, in which they change the solubility of the synthesized
peptide in halogenated solvents or THF. As a result, the obtained
peptide chain on the anchor support could be isolated easily and
efficiently by simple precipitation and washing using methanol
or acetonitrile, and the peptide with the anchor support molecule
at the C-terminus was obtained in an 86% yield. However, after
final deprotection of the obtained crude compound, we found a
contaminating peptide that lacked two amino acids at the C-termi-
nus. The observation of this deletion peptide indicated that the an-
chor support molecule 1 was released at the second amino acid
residue, due to diketopiperazine formation, and this was followed
by loading of an Fmoc-protected third amino acid (Fmoc-Ala-OH)
on 1 (Fig. 2).8
Diketopiperazine formation is a well-known side reaction that
occurs at the C-terminus. This side reaction is particularly favored
in peptide synthesis using Fmoc chemistry during the base-
induced deprotection of an Fmoc group at the dipeptide stage as
well as the coupling reaction of a third amino acid. In SPPS, steri-
cally hindered 2-chlorotrityl resin is frequently used as a support
linker to minimize diketopiperazine formation.9 When this chem-
istry was applied to the liquid phase, trityl-type compounds 3a and
3b were reported as anchor support molecules.10
The LPPS was next examined using the support molecules 3a
and 3b to evaluate the amount of diketopiperazine formation.
However, their amino acid esters readily degraded in a homoge-
neous solution of methanol, probably due to the electron-donating
character of alkoxy groups, although esterification proceeded to
give the product in good to excellent yields (Table 1). Thus,
modification was made through the introduction of stronger elec-
tron-withdrawing groups in the two phenyl rings of 3a and 3b
compared to the chlorine atoms at the 4-position, to achieve suffi-
cient chemical stability. The number of electron-donating alkoxy
groups was also investigated. These designed anchor compounds
(3a–e) were synthesized using gallic acid derivatives 4 as starting
materials, which were treated with Grignard reagents followed by
chlorination or bromination (Table 1). The loading of protected
amino acids was accomplished by treatment with the correspond-
ing amino acid in the presence of iPr2EtN.
The derivative with three alkoxy chains and two fluorine atoms
3c showed a 95% yield in the esterification step, but the resulting
ester showed insufficient stability in methanol at 40 °C and was
cleaved in 10% AcOH/CHCl3, indicating that this ester bond is easily
degraded under weakly acidic conditions, similar to that of HOBt.
Therefore, this compound is not suitable for use as an anchor sup-
port molecule in our LPPS. A compound with two alkoxy groups 3d
was also tested, and was associated with a severe reduction of
reactivity in the esterification step. Its brominated derivative 3e
was synthesized and tested with the expectation of increased reac-
tivity for loading of the first amino acid residue. However, com-
pound 3e also showed negligible reactivity in ester bond
formation. These results indicated that it is difficult to achieve a
sufficient balance between the chemical stability of the ester bond
and a high loading yield with a protected amino acid using the tri-
tyl-type anchor support molecules 3a–e.
Fluorene derivatives are sterically hindered similar to those
with a trityl group, and thus they are used as protecting groups
for carboxylic acids in the side-chain of amino acid residues.11
We next applied this chemistry to our LPPS, and used fluorene-type
support compounds possessing long alkoxy chains to reduce dike-
topiperazine formation (Table 2). Generally, the alkoxy chains pos-
sessing more than about 30 carbon atoms in total are sufficient for
the elongation of peptide sequence using our LPPS protocol. 2-Hy-
droxy-fluorenone 7a or 2,7-dihydroxy-fluorenone 7b was stirred
with alkyl bromide under basic conditions, followed by a Grignard
reaction to afford tertiary alcohol 9. Treatment of 9 with acetyl bro-
mide provided fluorene-derived anchor support compounds 10.
The protected amino acids were loaded on this support molecule
in the presence of iPr2EtN.
For compound 10a, which possesses an unsubstituted phenyl
ring and two alkoxy chains, the yield of esterification was 96%,
and the obtained ester was more stable than the trityl-type support
molecules in both methanol and 10% AcOH/CHCl3 (the observed
degradations were 1.8% and 3.2%, respectively). The introduction
of a chlorine atom 10b or fluorine atom 10c was also associated with
an excellent loading yield and sufficient stability of the ester bond.
The trifluoromethyl derivative 10d showed an 88% yield in the
esterification step with no sign of degradation. The alkoxy chains
were subjected to further optimization. The 2-docosyl fluorene-
type anchor molecule with chlorine atom 10e showed an excellent
yield for amino acid loading with less than 1% cleavage of the ester
bond. Compound 10f, which has a 2-[12-(docosyloxy)dodecyloxy]-
type chain and a fluorine atom, showed the best result among the
anchor molecules tested, with a 95% loading yield and no sign of
ester bond cleavage. The introduction of trifluoromethyl groups
10g, which have a stronger electron-withdrawing effect than fluo-
rine atom, showed reduced reactivity in the esterification step.
These results indicated that fluorene-type support molecules 10c–
g could provide sufficient chemical reactivity in the loading of a
protected amino acid, and efficient stability under the homoge-
neous condition used for the LPPS protocol. Moreover, on the anchor
support molecule 10f, Fmoc-protected amino acids could be selec-
tively detached with other protecting groups intact under weakly
acidic conditions, such as in 2%TFA/CHCl3 or hexafluoroisopropanol
(HFIP) (Fig. S1 in Supplementary data). These results suggest that
this LPPS method with the anchor support compound 10c–g can
give a protected peptide acid which has a free carboxyl group for
further fragment condensation reactions.
mAU
220nm,4nm (1.00)
500
Fmoc-Tyr-Phe-Ser-
Ala-Pro-Gly-OH
Fmoc-Tyr-Phe-Ser-Ala-OH
250
0
min
25.0
10.0
12.5
15.0
17.5
20.0
22.5
Figure 2. HPLC chart of the obtained crude peptide Fmoc-Tyr-Phe-Ser-Ala-Pro-Gly-
OH, synthesized by LPPS using compound 1 as an anchor support compound at the
C-terminus.
We next treated the Fmoc-protected dipeptide Fmoc-Pro-Gly on
two types of anchor support molecules 1 and 10f under basic

1938
D. Takahashi, T. Yamamoto / Tetrahedron Letters 53 (2012) 1936–1939
Table 1
Evaluation of the loading efficacy and stability of the ester bond on trityl-type anchor compounds
2~5 eq
X
X
X
BrMg
O
Z-Ala
SOCl
X
2
R
R
HO
O
1
1
Y
R
R
Z-Ala-OH
i
R
R
1
2
R
R
MeO
or AcBr
1
1
2
Pr EtN
2
o
CHCl
3
THF, 50 C
2
R
1
R
R
R
1
1
1
X
X
X
4
5
3
6
Anchor support compound
R1
R2
R3
X
Y
Loading yielda (%) (Z-Ala esterification)
Degradation of 6b (%)
MeOH (40 °C)
10% AcOH/CHCl3
3ac
3bc
3cd
3dd
3ed
OC18H37
OC22H45
OC18H37
OC22H45
OC22H45
OC18H37
H
OC18H37
H
H
OC18H37
OC22H45
OC18H37
OC22H45
OC22H45
4-Cl
4-Cl
3,5-F
3,5-F
Cl
Cl
Cl
Cl
Br
97
89
95
0e
69
19
4
100
74
63
Not tested
Not tested
Not tested
Not tested
3,5-CF3
0e
a
Loading efficacy was determined by the isolated yield of corresponding compound 6 after precipitation and washing with acetonitrile. Based on the results of NMR
spectroscopy, the isolated compound 6 had at least 95 % purity.
b
Degradation of the ester bond between Z-Ala and anchor support compounds was determined at 1 h by measuring the released Z-Ala-OH by HPLC analysis.8
Ref. 10.
This work.
50 °C.
c
d
e
Table 2
Evaluation of loading efficacy and stability of ester bond on fluorene-type anchor compounds
(2 ~ 5 eq)
X
X
BrMg
O
O
alkyl bromide,
K2CO3 (1.5 eq)
DMF, 80oC
HO
R
R2
R2
OH
R1
R1
THF, 50oC
8
7a: R = H
9
X
7b: R = OH
X
Z-Ala-OH (2 eq),
iPr2EtN
Z-Ala
Br
O
AcBr (5 eq)
R2
R2
R1
CHCl3, r.t., 3 h
CHCl3, r.t. ~ 50oC
R1
10
11
Anchor support compound
R1
R2
X
Loading yielda (%) (Z-Ala esterification)
Degradation of 11b (%)
MeOH (40 °C)
10% AcOH/CHCl3
10a
10b
10c
10d
10e
10f
OC22H45
OC22H45
OC22H45
OC22H45
OC12H15 OC22H45
OC12H15 OC22H45
OC12H15 OC22H45
OC22H45
OC22H45
OC22H45
OC22H45
H
H
4-Cl
3-F
3-CF3
4-Cl
3-F
96
98
93
88
92
95
85
1.8
1.0
0.3
3.2
3.0
0.3
Not detected
0.4
Not detected
Not tested
Not detected
0.8
Not detected
Not tested
H
H
10g
3-CF3
a
Loading yield was determined by the isolation of corresponding compound 11 after precipitation and washing with acetonitrile. Based on the results of NMR spec-
troscopy, the purities of the obtained compounds 11 were at least 95%.
b
Degradation of the ester bond between Z-Ala and the anchor support compound was determined at 2 h by measuring the released Z-Ala-OH by HPLC analysis.8
conditions to remove the Fmoc group, to evaluate the rate of
diketopiperazine formation. The release of fluorene alcohol 9f
was slower than that with benzyl-type compound 1, and only
1.4% of 9f was released after treatment for 6 h with an excess
amount of Et2NH (15 equiv to 10f)/DBU (1.5 equiv) in chloroform
(details are provided in Supplementary data). This result clearly
indicated that 10f induces a minimum amount of diketopiperazine
formation under normal conditions for Fmoc deprotection (Et2NH/
DBU in chloroform for 1 h), which is a significant advantage over
benzyl-type anchor supports, such as compound 1. Furthermore,
the fluorene alcohol 9f did not react with a protected amino acid
in the presence of EDCÁHCl and HOBt in chloroform. Thus, the
deletion peptide that lacks two C-terminal amino acid residues
was not observed. These results suggested that the fluorene-type
compound 10f is an efficient and useful anchor support compound
for LPPS.
Elongation of a peptide sequence was performed on anchor
support compound 10f using Fmoc-chemistry to synthesize a
protected peptide acid which has a free carboxyl group at the C-ter-
minal with protecting group-intact Fmoc-Tyr(tBu)-Phe-Ser(tBu)-
Ala-Pro-Gly-OH 12 (Scheme 1). Fmoc-Gly-OH was introduced to
10f in the presence of iPr2EtN in chloroform to afford amino acid
ester 13. After precipitation using a polar solvent,12 deprotection
of the Fmoc group was carried out using a mixed solution of DBU
and Et2NH in chloroform.13 The N-terminal free amino acid with
the anchor 14 was precipitated and washed with acetonitrile, and

D. Takahashi, T. Yamamoto / Tetrahedron Letters 53 (2012) 1936–1939
1939
1) Fmoc-Pro-OH,
EDC.HCl, HOBt,
CHCl3
1) Fmoc-Gly-OH,
iPr2EtN, CHCl3
2) precipitation
(MeCN)
F
1) Et2NH, DBU
F
F
CHCl3
2) precipitation
(MeOH)
2) precipitation
Fmoc-Gly
H-Gly
(MeCN)
Br
O
O
O
OC22H45
12
O
OC22H45
12
O
OC22H45
12
10f
13
14
F
F
Fmoc-Tyr(tBu)-Phe-Ser(tBu)-Ala-Pro-Gly
O
Pepetide chain
elongation
Fmoc-Pro-Gly
2%TFA
CHCl3
O
Fmoc-Tyr(tBu)-Phe-Ser(tBu)-Ala-Pro-Gly-OH
O
OC22H45
12
C22H45
O
O
12
12
15
yield from 16 =98%
16
Total elongation yield 83%(11steps)
purity 97%
Scheme 1. Synthesis of protected peptide with a free C-terminus using the fluorene-type anchor molecule 10f.
mAU
500
400
300
200
220nm,4nm (1.00)
Supplementary data
The synthetic procedure of fluorene-based anchor molecules,
general procedure for peptide synthesis using the anchor support
molecule 10f, the release of Fmoc-Ser(t-Bu)-OH from the anchor
support compound 9f, and the rate of diketopiperazine formation
determined by the release of anchor support compound 9f or 1.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.tetlet.2012.02.006.
100
0
10.0
12.5
15.5
17.5
20.0
22.5
25.0
27.5
min
Figure 3. HPLC chart of protected peptide acid Fmoc-Tyr(tBu)-Phe-Ser(tBu)-Ala-
Pro-Gly-OH 12 synthesized with our LPPS using compound 10f as an anchor
support compound at the C-terminus.
References and notes
coupling of next Fmoc-amino acid was then performed using
EDCÁHCl and HOBt as coupling reagents. When these coupling-pre-
cipitation–deprotection–precipitation processes were repeated, the
target fully-protected peptide 16 was synthesized on the anchor
molecule. After the anchor support molecule was selectively
detached using 2%TFA/CHCl3, the target protected peptide acid
Fmoc-Tyr(tBu)-Phe-Ser(tBu)-Ala-Pro-Gly-OH 12 was obtained in
an excellent total yield of 81% and 97% purity in 12 steps (Fig. 3).8
No sign of diketopiperazine formation was observed during elonga-
tion of the peptide chain.
}
}
1. (a) Kisfaludy, L.; Schon, I.; XSzirtes, T.; Nyéki, O.; Low, M. Tetrahedron Lett. 1974,
15, 1785; (b) Carpino, L. A.; Ghassemi, S.; Ionescu, D.; Ismail, M.; Sadat-Aalaee,
D.; Truran, G.; Mansour, E. M.; Siwruk, G. A.; Eynon, J. S.; Morgan, B. Org. Process
Res. Dev. 2003, 7, 28; (c) Eggen, I. F. Org. Process Res. Dev. 2005, 9, 98.
2. Pillai, V. N. R.; Mutter, M.; Bayer, E.; Gatfield, I. J. Org. Chem. 1980, 45, 5364.
3. Isokawa, S.; Kobayashi, N.; Nagano, R.; Narita, M. Bull. Chem. Soc. Jpn. 1980, 53,
1028.
4. Mizuno, M.; Miura, T.; Goto, K.; Hosaka, D.; Inazu, T. Chem. Commun. 2003, 972.
5. Tamiaki, H.; Obata, T.; Azefu, Y.; Toma, K. Bull. Chem. Soc. Jpn. 2001, 74, 733.
6. (a) Tana, G.; Kitada, S.; Fujita, S.; Okada, Y.; Kim, S.; Chiba, K. Chem. Commun.
2010, 8219; (b) Chiba, K.; Kono, Y.; Kim, S.; Nishimoto, K.; Kitano, Y.; Tada, M.
Chem. Commun. 2002, 1766; (c) Kitada, S.; Fujita, S.; Okada, Y.; Kim, S.; Chiba, K.
Bioorg. Med. Chem. Lett. 2011, 21, 4476.
Based on an investigation of anchor support molecules, we
found that fluorene-based molecules 10c–g are useful in our LPPS.
These molecules were synthesized from fluorenone derivatives in a
good yield, and showed sufficient loading efficiency and stability of
the ester bond during elongation of the peptide sequence. With the
anchor support molecule 10f, the peptide chain was prepared by
the repetition of simple reactions and precipitation steps without
either the formation of diketopiperazine or the absence of protect-
ing groups in the side-chain. As a result, an efficient and easy-han-
dling LPPS protocol (AJIPHASEÒ) was successfully developed to
prepare protected peptide acids, which are useful in fragment
condensation to obtain peptides with diverse sequences and chain
lengths. The synthesis of bioactive peptides using the AJIPAHSEÒ
protocol is in progress, and will be reported elsewhere.14
7. Hirose, T.; Kasai, T.; Akimoto, T.; Endo, A.; Sugawara, A.; Nagasawa, K.; Shiomi,
K.; Omura, S.; Sunazuka, T. Tetrahedron 2011, 67, 6633.
8. The obtainedpeptide was analyzed using thefollowing HPLCsystem: SHIMADZU
LC20A; C-18 reversed-phase column: YMC-Pack ODS-A, 150 mm  4.6 mm ID
(YMC Co., Ltd, Kyoto, Japan); solvent: 0.05 % TFA/acetonitrile = 80:20 to 20:80 in
25 min; flow rate: 1.0 mL/min; detection: 220 nm.
9. (a) Barlos, K.; Gatosa, D.; Kallitsisa, J.; Papaphotiua, G.; Sotiriua, P.; Wenqinga,
Y.; Schäfer, W. Tetrahedron Lett. 1989, 30, 3943; (b) Barlos, K.; Gatos, D.;
Kapolos, S.; Papaphotiou, G.; Schafer, W.; Wenqing, Y. Tetrahedron Lett. 1989,
30, 3947.
10. Chiba, K.; Kim, S.; Kono, Y. Patent Application WO2007/122847, 2007.
11. (a) Litor, B.; Orlich, S.; Kunz, H. Synlett 1999, 1136; (b) Mergler, M.; Dick, F.; Sax,
B.; Weiler, P.; Vorherr, T. J. Pept. Sci. 2003, 9, 36.
12. We used acetonitrile or methanol as solvents to precipitate the targeted
peptides. Generally, methanol is better solvent for removal of the residual
amino acid compared to acetonitrile, but has potential risk of methanolysis
during the deprotection of Fmoc group under the basic condition.
13. Under this deprotection condition, negligible amount of aspartamide
formation was observed during the synthesis of peptide sequence possessing
Asp at the second position (Takahashi, D., unpublished result).
Acknowledgement
14. Takahashi, D., manuscript in preparation.
We would like to thank Mr. Tatsuya Yano and Dr. Tsuyoshi Izu-
hara for their help with the experiments.